Dean Edwards (PhD '77), Engineer, Roboticist, and Leader in Electric Vehicle Research

Long before there was Tesla, General Motors developed the EV1, the world's first commercially available electric vehicle. Drivers loved it, but soon GM infamously ended the program, thereby delaying the widespread adoption of electric vehicles by at least a decade. In the discussion below, Dean Edwards provides an authoritative account detailing the Jet Propulsion Laboratory's crucial contributions to EV technology, in partnership with the Department of Energy, which sought to develop alternatives to internal combustion automobiles. Edwards conducted pathbreaking work on battery technology before lithium ion batteries became the standard for EVs. His contributions in sealed lead acid batteries remain a standard energy supply for numerous applications.

With the end of the DOE-JPL partnership, Edwards moved to the University of Idaho where, over the past thirty five years, he concentrated on robotics, with specific application for forestry. Edwards remained connected to EVs with his leadership on a college team to build a hybrid electric prototype, and in retirement, he has focused on developing autonomous underwater vehicles. At the end of the discussion, Edwards reflects on the research culture and can-do spirit that infuses Caltech and JPL, and why these qualities have made Pasadena central to the early history of electric vehicles.

DAVID ZIERLER: This is David Zierler, Director of the Caltech Heritage Project. It is Tuesday, August 30, 2022. I am very happy to be here with Professor Dean Edwards. Dean, it's great to be with you. Thank you for joining me today.

DEAN EDWARDS: I'm glad to be here.

ZIERLER: To start, would you tell me your title and institutional affiliation?

EDWARDS: At this moment, I'm Professor of Chemical Engineering, Emeritus, at the University of Idaho. I started at UI in mechanical engineering, but because of my battery research and other interests, I eventually switched over to chemical engineering, although I had a joint appointment for a few years. I was both a Professor in Mechanical Engineering and Chemical Engineering, and now Professor, Emeritus, in Chemical Engineering.

ZIERLER: When did you go emeritus?

EDWARDS: I went half-time for a while. I fully retired when I was 70, which would've been in 2019, just before COVID.

ZIERLER: That's good timing.

EDWARDS: Yeah.

ZIERLER: Are you enjoying a true retirement, or are you active at all in research?

EDWARDS: Before I became Professor Emeritus, I was part-time, just doing research for a few years. I'm still involved with some undergraduates and one graduate student, so I kind of have a hand in research, even today. Maybe half a day a week or so.

ZIERLER: Between your affiliations and your academic training, at the end of the day, what kind of engineer are you, if you had to pick?

EDWARDS: I would say I'm a systems engineer. I was interested in control theory, I did nonlinear analysis, fuzzy logic control systems. My background is more in systems and controls. In my Ph.D. thesis at Caltech, I used discrete, time domain analysis to determine both the local and global stability of control systems for switching regulators, which use power electronics. This thesis topic introduced me to power electronics and the use of switching to improve the efficiency of power conversion. When I went to JPL, I got involved with EV power trains, and that led to batteries, and that's how I ended up with that kind of background. Mainly, I'm a control systems person, which is ubiquitous to engineering.

I've taught control classes in Mechanical Engineering, Electrical Engineering, and Chemical Engineering. The ME control class was an elective which was very popular and so I would teach that class once a year. The EE controls class was a required class for EE students and, because the courses were similar, they could substitute the ME controls class to fulfill the EE requirement. Chemical Engineering required two semesters of control courses which were oriented more towards process controls. Because control theory is used to analyze feedback systems, control analysis is used in all kinds of fields other than engineering including economics and biology, to name a few.

I enjoyed control systems and that is my academic area, but I was very interested in a career in robotics. In hindsight, choosing robotics as a career when microprocessor technology was not readily available is now surprising to me. I guess I was optimistic about computer technology but when I graduated with a BS in Mechanical Engineering in 1972, I remember being enamored with the four-function calculator. In 1977, when I received my Ph.D. from Caltech, the personal computer was just being introduced and that technology would provide the control complexity required for robotics.

ZIERLER: After JPL, did you spend your entire career in Idaho?

EDWARDS: Yeah, I went from JPL to Idaho in 1986, and I've been here ever since.

ZIERLER: Tell me generally about engineering in Idaho. What are some of the big areas of impact there?

EDWARDS: Well, it's the land-grant institution for the state of Idaho, and is the primary research institution in the state. Because of its land-grant status, UI receives a lot of funding from USDA (US Department of Agriculture). We have all the engineering programs – mechanical, electrical, computer engineering, computer science, we have a chemical engineering program, and civil engineering. We also have some environmental engineering. We have the full complement of engineering programs at Idaho. Some notable areas of research in COE (College of Engineering) included the Microelectronic Research and Communications Institute (MRCI), power electronics, Center of Intelligence System Research (CISR), Center for Advanced Thermodynamic Studies, material engineering (including nuclear), computer security, and environmental engineering.

JPL and Electric Vehicles

ZIERLER: What have been some of the major research projects you've undertaken in your career?

EDWARDS: The work I did on electric vehicles starting at JPL was my most important research project and the one on which I spent the largest part of my professional career, both at JPL and UI. When I first started working at JPL, right after my Ph.D., I soon found myself more interested in working in the energy area than in the space program.

ZIERLER: Were you aware of how DOE became involved in supporting JPL, what the specific initiative was?

EDWARDS: Fortunately for me, about the time I started working at JPL, NASA funding was being reduced and the NASA labs started to take on other projects, mainly Department of Energy projects. That's how I got involved with electric vehicle research which included battery and power train research, which I did with Wally Rippel. We had a big program with EPRI (Electric Power Research Institute). Before that, we received some funding from DOE.

NASA and DOE had some sort of letter of understanding. I think initially, JPL had programs for testing electric vehicles, and DOE gave them money for that. But they had, I think, a letter of understanding that JPL could support Department of Energy projects. A lot of that work was done off campus down in the Foothill Boulevard area. They were doing solar cell work down there and electric vehicle work. It was moved off the JPL campus. Although, we reported to the same people back at the main lab that we did before.

ZIERLER: Did you express interest in switching departments? Were you pulled into it? How did the transfer happen?

EDWARDS: I found out about the DOE projects, and I'd go talk to the project manager who I wanted to work for. JPL is a matrix management organization. You work in an area or specialty like power electronics or control systems, but you also work on projects which need your skills for that project. You may have a spacecraft that needs somebody to build a switching regulator, so then you'd go to the line management department that has people who work on power electronics. You'd get somebody from line management to do that switching regulator for your project.

If you're interested in a project, go talk to the project manager, they can talk to the line manager, and see if they can get some of your time. That's kind of how JPL worked. I didn't really seem to have too much trouble getting that done. But actually, we ended up with EPRI as the major sponsor for our EV research, and we developed that funding source ourselves. And it was in the energy area that we were interested in, so we were able to work with EPRI, even though they weren't DOE or NASA.

ZIERLER: When did you first connect with Wally Rippel?

EDWARDS: It was at JPL. I think it was probably a year or two after I started there. It was when the energy programs grew. Wally came in, and he knew the person in charge of the battery program at JPL. That's kind of where he started working. I was working testing EVs and had moved back to working in the power electronics department. We had mutual interests and started meeting for coffee and became good friends.

We talked a lot about EVs and EV research. When we first started meeting for coffee, DOE was funding a lot of different battery research projects and we started joking about the "battery of the month club." All these batteries had high energy densities but did not appear to be very viable for EV applications because of other problems such as cost, safety, life, recyclability, etc. The problem with DOE was that they were trying to develop a battery rather than an electric vehicle. The type of EV being designed would influence the type of battery selected for that design. People were mainly interested in a 100-mile range EV, and they argued that if "we can land a man on the moon, why can't we develop a 100-mile EV." In our EV discussions over coffee, we decided that if you do everything right in an EV design, that you could theoretically use a lead acid battery for a 100-mile EV. The specific energy of a lead acid battery would only need to be about 30% higher than conventional lead acid batteries. In addition, we believed that the lead acid battery would need to be a maintenance free, sealed design and have a high-power density. Our thinking was "if possible, why shouldn't you develop a lead-acid battery for a limited range EV?"

ZIERLER: Did you work with him directly on batteries?

EDWARDS: Yeah, we had a number of patents on batteries and power electronics. We started looking at lead-acid battery designs that would have improved energy density, high power density, and be sealed for low maintenance. We developed a bipolar battery that achieved these design goals and had very high-power density. Unfortunately, the battery had leakage problems which caused us to develop an alternative design which we called a "quasi-bipolar battery." This new battery had acceptable power density but had heat transfer and manufacturing issues.

Our design iterations with high-power batteries gave us insight into how to achieve high power densities with different battery topologies. Because we wanted a sealed battery, a horizontal plate topology was possible. Most lead acid batteries are not sealed and use vertical plates to allow for gases generated during charging to escape. A sealed battery allowed for a horizontal plate topology where multiple connectors could be used to provide high power, heat transfer from the cells could be greatly improved, and uniform compression could be applied to the plates for long life. This horizontal plate battery could achieve the energy and power density performance needed for this application and do so in a sealed design. This was the design on which we settled.

We were not able to secure funding from DOE for the horizontal plate battery. In fact, we had a proposal to DOE on the quasi-bipolar battery which my JPL manager thought would be funded. He did not believe the horizontal plate battery would be funded because its design was more conventional. Although we recognized this problem, the horizontal plate, sealed lead-acid battery was the most attractive battery for our limited range EV, and we did not want to waste our time on other batteries. I pitched the horizontal plate battery to DOE, but they were not interested.

This was a recurring problem with DOE as they were not interested in funding lead acid batteries even if they were the best option for an EV. They were interested in battery research but not in the context of developing a battery for a practical electric vehicle. Fortunately, the Electric Power Research Institute (EPRI) was interested in developing the appropriate technology for EVs and did fund our horizontal plate battery.

With EPRI funding, we worked with two battery companies, Eagle-Picher and Concord Battery, to implement our horizontal plate design with their manufacturing capabilities. The major issue EPRI had with our battery development was the sealed design. Independent battery consultants used by EPRI were skeptical that a sealed, lead acid battery could achieve the cycle life required for EV applications. EPRI did fund JPL to do some preliminary work with the two battery companies to fabricate and test some of our cells. These cells were cycled and showed promise for attaining the desired cycle life. These preliminary results encouraged EPRI to increase the funding for our battery development.

ZIERLER: Besides batteries, did you work with him on other EV technologies?

EDWARDS: Once we had developed a good relationship with EPRI through our battery development program, we were able to interest EPRI in our AC drive development that would complement our high-power battery. An AC drive system had more than two times the power density of a DC system. The high torque at low rpm for the AC system also provided tremendous acceleration for an EV and became an important feature for EV acceptance. The EPRI AC powertrain consisted of our high power, horizontal plate sealed lead acid battery and an AC drive. However, all our funding was contingent on demonstrating adequate life for the battery.

The problem with implementing an AC drive was related to the electronic controller (i.e., inverter) needed for the AC motor. The power semiconductors needed for use in the inverter did not exist. To overcome this limitation, Wally and I investigated what we called hybrid topologies which consisted of using a combination of different semiconductors to achieve the high-power switching needed for this application. Two of these topologies consisted of using field effect transistors (FETs) with either bi-junction transistors (BJTs) or ASCRs (asymmetrical silicon-controlled rectifiers). We also developed a baseline inverter with high power BJTs where more conventional drive circuitry was used.

The original funding for the inverter work started in 1981 with a Caltech President's Fund to demonstrate the feasibility of these hybrid topology designs. Using the results of this work, we were able to persuade DOE to continue our inverter development. A final report was provided to DOE in November 1984 that documents the evaluation of these topologies and the testing of a 60 kw BJT inverter. The drive circuitry used for controlling the BJTs consisted of discrete devices. Eventually (i.e., early 1990s), the semiconductor of choice became the IGBT, integrated-gate bi-junction transistors, where the drive components were integrated onto the same chip as the power BJT. I believe the EV1 inverter used IGBTs as did the AC Propulsion inverters.

There were also power MOSFETs (metal-oxide semiconductors) available when we were first looking at developing AC drives. These devices could handle the needed EV voltages and had many desirable characteristics but could not handle the large currents needed for the application. Wally talked me out of using them because of the current limitation which would have necessitated using hundreds of these devices and connecting many in parallel. However, for the Impact vehicle, Alan Cocconi did use power MOSFETs for the inverter.

ZIERLER: The patents you developed with Wally; this was all within the context of JPL? Or these were outside ventures?

EDWARDS: No, these were all through JPL. JPL was a NASA lab, and we would file what we thought might be patentable ideas at JPL. Many times, these ideas would be turned into NASA Tech Briefs even when the patent attorneys did not think a patent application was warranted. We were awarded patents for bipolar, quasi-bipolar, and horizontal plate batteries. We also had other battery patents including methods for applying compression to the face of lead acid battery plates to improve battery life and lead coated glass fibers used to provide conductivity in the positive plates. We also had patents on the work we were doing with inverters including the two hybrid topology inverters previously discussed. These are the patents I can remember but I believe we did have some more. Although NASA, through JPL, developed the Tech Briefs and filed for the patents, all these patents belonged to Caltech, who managed JPL.

ZIERLER: What kinds of applications were you thinking of at the time? In other words, was the concept of a commercially available EV one of the end uses as you were working on this?

EDWARDS: That was probably about the only end use we had in mind. However, we did have some interest in developing inverters for industrial drives. That application for our inverter technology probably made more commercial sense than the EV application, if we'd thought about it. Our inverter technology was being developed strictly for electric vehicles but would have had industrial application for controlling 60-Hertz induction motors in a much more efficient manner. There are a lot of inverters out there now that you can use in industry and for home use, too.

ZIERLER: The reason I asked about end uses was, I wasn't aware that JPL was thinking about supporting research that would go into an electric vehicle. I thought it might've been for some space application, perhaps.

EDWARDS: No, we were strictly in the energy area, working down at Foothill. That was what we were interested in doing. JPL always made the argument that "This is kind of generic technology, so it could be applied to space applications," but our systems could not be applied directly to space applications. I don't believe JPL would have ever allowed this research if they, along with the other NASA centers, had not been under financial stress from the lack of funding from NASA. These energy programs began to be shed as soon as the NASA funding started to return.

Industry Considerations

ZIERLER: Did you ever talk with Wally about going into business or industry?

EDWARDS: We did talk towards the end of my tenure at JPL about developing industrial inverters to make ac motors more efficient. For example, a pump attached to an induction motor will run at the speed associated with the 60 Hz line frequency. When the motor and pump are operated full-on, then the efficiency is usually high but, at lower flow rates, the pump is usually throttled which can result in large losses. Inverters can control the speed of the motor and pump so that throttling is unnecessary. Although the inverter may be expensive, the energy savings from more efficient use of the motor may justify the purchase of an inverter.

Wally and I did have some conversations about starting a business producing these inverters. We even approached some people about supporting this venture and they probably would've provided support, to some extent, but we didn't know enough about industrial inverters and how to market and sell them. Obviously, there's a big market for it now. If I had stayed at JPL, we probably would have tried to start a company for producing these inverters.

ZIERLER: Do you have a sense of the degree to which DOE support made this research possible? In other words, would JPL have gotten into this area without DOE's interest?

EDWARDS: I don't think so. Again, JPL was a NASA center, and I don't think they would have ever become interested in energy and being funded by DOE if they didn't need the money. The DOE funds allowed JPL to keep their manpower and support their labs. I believe the arrangement benefitted JPL, NASA, and DOE but would not have happened if NASA hadn't lost funding.

ZIERLER: Tell me about the push and pull factors leading to your decision to leave JPL and join the faculty at Idaho.

EDWARDS: At the end of 1984, we had demonstrated the technology for both a high-power sealed lead acid battery and an AC drive. EPRI had provided funding for Eagle-Picher and Concorde to build battery cells based on our design (see Fig.1), and the test results with these cells were very encouraging. They were increasing funding to JPL to build a statistically significant number of cells for life testing. EPRI also agreed to fund a lab at JPL for doing life testing.

We had also developed, with funding from DOE, a 60-kw inverter for an AC drive (see Fig.2). Although both the battery and AC drive needed additional development, the feasibility was demonstrated, and the combination provided an AC powertrain for a limited range EV. EPRI picked up funding for our AC Drive development work from DOE and included it with the battery development. The battery and AC drive was the AC powertrain that we were developing for EPRI as described earlier.

Although we were having success with developing the battery and inverter technology for EVs, we realized that the importance of our work would not be understood unless it was demonstrated in a very efficient EV. In early 1985, we approached Paul MacCready at AeroVironment, Inc. about developing this very efficient vehicle where our sealed, lead acid battery and AC electric drive system could be demonstrated. Paul was interested and subsequently wrote the white paper for "The ElectroSpirit Program," published in August of 1985.

In the ElectroSpirit proposal, Paul delineated the responsibilities for the entities involved in the proposed work. JPL was responsible for developing an advanced electric powertrain consisting of the high performance, horizontal plate, sealed lead-acid battery, and an AC (alternating current) induction motor and controller. Because EPRI was funding JPL for the battery and the AC propulsion system development, EPRI would also have been part of the ElectroSpirit program. Paul also mentioned using a small trailer with an IC engine and generator for increasing the range of the EV when required. Unfortunately, the GM Advanced Concept Center was not willing to fund the ElectroSpirit program and EPRI was not comfortable in funding an entire EV development program by themselves.

If GM had funded the ElectroSpirit Program in 1985, I wouldn't have left JPL. I had both personal and professional reasons for leaving including wanting to return to Idaho. During my vacation that summer, I visited the University of Idaho and talked to some people about possible faculty positions. They had a position that was in my area and for which I applied after I learned the ElectroSpirit Program was not funded. I was not optimistic about receiving a job offer from UI but had decided that if I did, I would accept it. Although we were disappointed about the GM decision, we still had the EPRI powertrain program and were still committed to finishing that development.

Around this time, JPL made the decision to drop the energy programs they had pursued to help stabilize their funding after cuts were made to NASA in the 1970s. They pledged to continue the energy programs already funded but the EPRI work was problematic. The program was only funded through Fall 1986 and additional funding was contingent on achieving a required number of cycles on the test cells. In addition, I learned from discussions about the ElectroSpirit Program that EPRI's upper management was not as supportive of our work as was our program manager, Dave Douglas.

At the time of the ElectroSpirit proposal, EPRI was still concerned that a sealed design would not have the required cycle life. They wanted the horizontal plate, sealed lead acid battery to have good cycle life (i.e., 800 cycles). We had some initial cycle data that looked promising, but they wanted proof that a statistically significant number of test cells could achieve a minimum of at least 50-100 cycles before they would continue the development. They funded both the fabrication of the test cells and the JPL battery laboratory for testing these cells. In the Fall of 1985 and the Spring of 1986, we worked with Concorde and Eagle Picher to fabricate our design with their manufacturing equipment for these tests. We also installed at JPL the test equipment for cycling these cells and the temperature chambers needed for these tests.

On January 28, 1986, we had a meeting with people from EPRI, Concorde, and JPL to discuss the progress on cell fabrication and the cycle tests planned for these cells. The meeting started at 8:30am and we had hardly sat down when our JPL Section Manager interrupted and asked to use the conference room as the space shuttle Challenger had blown up. As we were waiting in the hallway to be taken to another conference room our program manager, Dave Douglas, informed me that he was retiring from EPRI and would be leaving that summer. I immediately knew that we were not receiving any more funding and our program was ending in the Fall even if the cell tests showed good results. I had just recently accepted a job offer from UI and informed him that I would also be leaving at the end of summer. I believe we both felt bad about the program being cancelled but we had no choice but to move on.

Concorde and Eagle Picher delivered the cells to JPL for the EPRI program, and we started testing them in early Spring of 1986. When I left JPL to go to the University of Idaho (UI) in August 1986, most of the cells had over 100 cycles and some cells had close to 200 cycles. These results should have meant that EPRI would continue funding the battery development, but I believe the decision to discontinue the funding had already been made before we even started the tests.

ZIERLER: Did you remain involved in this work from afar, from Idaho?

EDWARDS: When I departed JPL for UI, I wanted to focus my research on robotics, which had been my initial interest and reason for attending Caltech. In fact, I didn't intend to continue any of my EV research at UI, either batteries or inverters. After a year at UI, in the Fall of 1987, I received a call from the person at JPL responsible for conducting the EPRI cell tests.

When the EPRI funding had ended a year earlier, JPL had simply allowed the tests to continue as the automatic cell cycling equipment required little effort and the cells little maintenance. During the year, the cells were cycled an additional 350 times. The cells had achieved 550 to 600 deep cycles when the testing was terminated. They simply could not justify continuing the effort any longer after a year without funding. They offered to send me the data, which I gladly accepted. This data was included in a journal paper and in a report to EPRI.

I am still grateful to the people at JPL who continued the cell tests and provided me with the cycling data even after the funding had ended. This data showed our sealed lead acid battery could achieve the cycle life needed for EVs. I believe that was the first paper showing that sealed lead acid batteries could achieve long cycle life. In addition, these same JPL people persuaded EPRI to donate the cycling equipment and temperature chambers to UI so that I could continue doing research on batteries. Without this help, I probably would not have continued my battery research. The people at JPL were very competent and professional and I appreciate everything they did for me. They are a great asset to the Caltech community.

I also remained involved with both AeroVironment and AC Propulsion after moving to UI. I had a contract with GM to support AeroVironment in their PNGV (partnership for a New Generation Vehicle, 1993-94) program. UI participated in the 1992 HEV Challenge and we purchased one of the first five AC drive systems that AC Propulsion manufactured. Wally encouraged me to buy one of their systems which, due to a generous donor, UI was able to do. I also had other interactions with both AeroVironment and AC Propulsion that were informal but valuable to UI.

The Impact of the Impact EV

ZIERLER: How closely did you follow the Impact vehicle development after returning to Idaho?

EDWARDS: When I left JPL, I felt bad about abandoning the EV research that was important to me and for which I had dedicated almost ten years of my professional life, much of it working with Wally. However, without GM or some other entity willing to fund the entire vehicle development, I didn't believe the battery and AC powertrain technology would be appreciated. Although I viewed the rejection of the ElectroSpirit Program as the end of EV development, Wally had the opposite view. He thought the ElectroSpirit proposal brought together the critical technology needed to produce a practical EV and, more importantly, the key players that could accomplish this task. In hindsight, this is probably true although the process was not straightforward.

I did believe that if anybody could break loose the critical technology and demonstrate it in a practical EV, it would be Paul MacCready. Six months after I left JPL for UI, Paul was approached by GM to compete in a solar car race across Australia, which they eventually won in spectacular fashion in the Fall of 1987. Having garnered a lot of good publicity for GM, Paul was able to approach them again about funding the "ElectroSpirit Program" and was successful. My understanding was that this funding was not a "slam dunk" even after all the good publicity GM gained from the solar car race, but Paul persevered. At the beginning of 1988, three years after Wally and I first approached Paul, GM funded AeroVironment to develop a 100-mile range EV. I believe Alan Cocconi, who had done the electric drive for the solar car, suggested developing a high-performance race car instead of the four passenger EV in the original "ElectroSpirit" proposal. AeroVironment called this vehicle the "Impact."

The Impact vehicle used a sealed lead acid battery with an AC drive. The vehicle had a 120-mile range and excellent acceleration and handling performance. Alec Brooks, who I knew from Caltech, was the project manager for both the solar powered Sunraycer and Impact electric vehicle. Alan Cocconi did the AC drive including the inverter for the Impact and Wally Rippel was a consultant on the project. GM's battery company, Delco Remy, developed a sealed, lead acid battery for the vehicle. AeroVironment engineered and built the very efficient vehicle required for this program. The development of the Impact took two years and was introduced at the LA car show in 1990 to rave reviews. Against a lot of odds, they persevered and demonstrated a viable EV.

The Impact EV had an AC Drive system which was developed by Alan Cocconi with help from Wally Rippel for AeroVironment. The inverter used power FETs for the switching semiconductors. While at JPL, we had previously evaluated these devices and found that they could not handle the current required for this application. However, Alan was able to develop designs for using many of these devices in parallel to deliver the required high current. This was no simple feat and was a testament to his engineering skills. When I talked to Wally recently, he was surprised that he could still remember the part number for the power FET that they used.

The torque, particularly at low speeds, that the AC drive could generate was truly memorable. This performance was one of the main reasons for GM's decision in 1990 to produce the EV-1, which was the production version of the Impact. Alan, along with Wally, started AC Propulsion and developed a 100 kw, AC drive system that they began selling in 1992. The inverter for this system used IGBTs (i.e., integrated gate, bi-junction transistors) which could handle both the voltage and high currents required for their AC drives. Interestingly, while at JPL we evaluated BJTs for use as the power semiconductors in two inverter topologies that we evaluated for EPRI. Of course, our work looked at using discrete components for the drive circuitry while the IGBTs integrated these components onto the power semiconductors used for the power switching. AC Propulsion would use this same IGBT technology to produce a 150 kw (200 hp) inverter which provided even greater acceleration and performance for EVs.

ZIERLER: Was funding your lead-acid battery research ever a problem?

EDWARDS: Finding research funding is always a problem. JPL had done a study on batteries for EVs, before I left for UI in 1986, which concluded that the limited range EV was the most promising EV and the lead acid battery the best battery candidate for this EV. This study was done by Keith Hardy for DOE so that they knew about the importance of lead-acid battery research. One must wonder why an organization whose responsibility was to develop EV technology would be so opposed to funding the most promising battery candidate but that was the major problem in developing my research funding.

I realized that if I was going to be successful, I needed to publish papers on battery research and to make contacts with other researchers and funding agencies. The first battery research funding I received was from the International Lead Zinc Research Organization (ILZRO). I was surprised when representatives from ILZRO contacted me about funding my research as I had never approached them about money. They traveled to UI and established a fellowship to fund one of my graduate students. This funding supported my first Ph.D. graduate student and allowed me to start developing battery models and investigate different designs for high performance, sealed lead-acid batteries. Things just kind of happened from there.

The session chair at the first conference I attended (August 1988) as a UI employee worked for GM. He was very interested in my paper and provided funding for me to travel to GM and make a presentation on my work. I would continue this informal relationship with people at GM for many years. Through these GM contacts, I received an invitation to attend the kick-off meeting (1993) for the PNGV (partnership in the new generation vehicle) program.

UI paid for my travel, and I was able to skip some classes and attend this PNGV kick-off meeting. Al Gore was the driving force behind this program, and I remember meeting him at a reception held at the Blair House (VP residence). I was one of the few people from academia in attendance. I also ran into Paul MacCready and Alex Brooks at the meeting and had lunch with them. I believe it was only the first or second time I had seen them since moving to UI. Although I did not know this at the time, GM would receive a contract with the PNGV program and UI would be a subcontractor to AeroVironment on that program.

The other meeting in 1988 where I presented a battery paper was sponsored by the Society of Automotive Engineers (SAE). The EV design discussed in my paper was like the design in the "ElectroSpirit" proposal where the horizontal plate, sealed Pb-acid battery and 60 kw AC drive system were used to power a very efficient vehicle. I reduced the specific energy of the battery that was consistent with our test data so that the vehicle only had a 107-mile range. I was surprised by the response of some members of the audience who were skeptical of the 100-mile range estimate and seemed offended by it. However, they did not challenge me on any of my numbers, but still seemed skeptical. At that time, I wondered how they would react to the Impact EV that was already being designed and would be displayed in January 1990.

I also developed a battery proposal for DOE based on the sealed, horizontal plate battery. The program was called "Assessment of Battery Technologies for Electric Vehicles" and the batteries were evaluated for an EV van application. Out of all the batteries evaluated, the sealed, horizontal plate, Pb acid battery had the highest technical merit and had the lowest risk. However, the battery was not funded.

Gary Hendrickson was the Idaho National Energy Laboratory (INEL) manager for this program and somebody I knew. I had talked to Gary before I wrote the proposal and specifically asked him if DOE would fund a Pb-acid battery development program. From the information he had, he assured me that DOE would consider funding lead acid batteries along with more advanced batteries. Unfortunately, Pandit Patil, the DOE Battery R&D manager, decided that any Pb-acid battery development should be funded by industry and not DOE. He was only interested in funding advanced, high-risk batteries. I also found out that DOE felt that politically they could not exclude Pb-acid batteries from being considered. As I suspected, they had no intention of funding any of the Pb-acid proposals even though they were the most promising for this application.

After the Impact vehicle was a sensation at the LA auto show (1990), I expected that funding for Pb-acid battery research would be greatly improved. However, when DOE founded the Advanced Battery Consortium (ABC), GM insisted that Pb-acid battery research be excluded from being considered for funding. They thought they had proprietary rights on Pb-acid batteries that they did not want to share. So, as one of my colleagues noted, "they formed this research consortium and then eliminated the best candidate from that research."

However, I was successful in obtaining funding from many different sponsors. My research was primarily on sealed lead-acid batteries, and I worked with General Motors (GM), Department of Energy (DOE), Exide, GNB, Office of Naval Research (ONR), the International Lead/Zinc Research Organization (ILZRO), AeroVironment, Concorde Battery Co., Advanced Lead Acid Battery Consortium (ALABC), Electric Power Research Institute (EPRI), Daramic, Avista, and two small battery start-up companies. Battery research was an important part of my research program at UI and an area where I am still doing some work.

Focus on Battery Research

ZIERLER: What kind of battery research did you do?

EDWARDS: The two physical processes that we modeled and that can limit the capacity of lead-acid batteries are diffusion and conductivity. For a reaction to occur in the battery electrode, a sulfate ion needs to diffuse to the point of reaction and an electron needs to be conducted away from the reaction. At high rates of discharge, the diffusion of ions into the electrode is usually what limits the reaction. At low rates, the conductivity can limit the reaction. Because the reaction changes the conductivity of the electrode, the electrode that is initially conductive becomes less conductive as lead sulfate forms, which is very nonconductive. From conductivity limitations, the maximum theoretical utilization of the active material in an electrode is about 60% (i.e., 50-55 % for positive plates and 60% for negative plates). About half the active material in lead-acid batteries cannot react due to this conductivity problem which is a big hit on the amount of energy these batteries can store.

In our research, we developed diffusion models and conductivity models for lead acid batteries. We were able to combine these models into one general model so that both the diffusion and conductivity processes could be analyzed. We used these models to evaluate different battery designs and different paste additives. Some of the paste additives we evaluated included different size particles, both conductive and nonconductive. We investigated the use of hollow, glass microspheres (HGM) in battery paste. This work persuaded Exide to fund our work on porous, hollow, glass microspheres (PHGM) where Savannah River National Laboratory (SRNL) fabricated the PHGMs. We have also evaluated ceramic fiber, both non-conductive and conductive.

In much of this work, I collaborated with Professor of Chemistry, I. Frank Cheng. A major problem we faced in our work was finding a conductive material that was also stable at the positive potential of lead acid batteries. We had very little success until Frank, with some of his students, accidentally discovered a carbon material that was conductive and stable at the positive potential of lead acid batteries. We were able to coat this carbon onto crystalline surfaces and had success with coating our PHGMs and ceramic fibers with this material, which I will call UI conductive carbon.

The work we did on the PHGMs for Exide showed that they could improve battery performance at higher discharge rates. However, the performance improvement was reduced because the PHGMs were non-conductive. Subsequently, we demonstrated that we could apply the UI conductive carbon coating to PHGMs so that the electrodes could possibly achieve their full predicted performance. Unfortunately, after the Exide program, we only had a small amount of PHGMs from SRNL. We discovered how to coat these PHGMs with the conductive carbon but did not have enough PHGMs to do plate experiments.

Research on PHGMs having conductive carbon coating looks very promising but without a good source of PHGMs to coat, the additives cannot be fabricated. At one time, the people at SRNL thought they would be able to surplus the equipment they used for fabricating the PHGMs to UI. Unfortunately, SRNL decided to warehouse rather than surplus the equipment capable of producing the PHGMs, so UI was not able to acquire the equipment. Although this development cannot proceed until a source of PHGMs can be found, carbon conductive coating technology has been demonstrated for this additive and the next step would be testing electrodes with this additive in cells.

The ceramic fibers having this conductive coating were found to help the plate's structure and life. These conductive fibers also improved utilization of the active material but not as much as predicted by our models. We are investigating the interface between the carbon and active material as a reason the utilization is not as high as the model predicts. These additives, when used in properly designed batteries, could provide high energy density at high power rates. This is still a research area of interest and one where I am active.

The initial discovery of the UI conductive carbon was the result of experiments performed on sand tar in crucibles. The sand tar residue at the bottom of the ceramic crucible that appeared after heating the crucible was a silver, metallic like substance that was conductive. This substance was both conductive and stable at the electrochemical potential of the lead acid battery positive electrode. The only substance we were ever able to identify where these two properties existed together. Because of the limited amount of material that could be fabricated with the crucible method, we needed six months to produce enough material for testing just a few electrodes having additives coated with this conductive carbon. However, the experiments that we performed verified that the carbon was stable in the positive plates of lead acid batteries.

We eventually discovered that additives in ceramic boats could be coated with this conductive carbon coating while heated in a tube furnace. Although many different sources of carbon can be used to create this conductive coating, we presently use diesel and nitrogen feeds for this purpose. We have also started to use a rotary kiln with this feed system to provide a uniform coating on the additive and to prevent a loss of product with the exhaust gases. The amount of conductive additive needed for the cell tests can now be easily produced and the fabrication methods are even amenable to being scaled to small production quantities.

Our present research is focused on establishing a good interface between the positive active material and the UI conductive carbon coating on the additive. For our purposes, we are only investigating the ceramic fibers but could also coat the PHGMs if they were available. We established that by heat treating the conductive coating on the additive that the coating becomes hydrophilic. Without the heat treatment, the UI carbon coating is hydrophobic and does not appear to bond very well with the active material. Mechanical compression could also be used to help with this interface. Although we have a good understanding of the problems and how to fix them, research funds for additive and plate fabrication and testing are not readily available and so progress is slow.

ZIERLER: Did you do any more work on your horizontal plate, sealed lead acid battery after moving to UI?

EDWARDS: In addition to our fundamental battery research, we also continued the development of the horizontal plate battery design for EV or HEV use. In one case, we participated in the General Motors PNGV (Partnership in a New Generation of Vehicles). In the other case we investigated using the battery in a military HUMVEE. The Humvee work was sponsored by ONR (Office of Naval Research). In both these cases, we worked with people at AeroVironment.

The UI was a subcontractor in GM's PNGV program. The goal of the program was to produce a diesel electric HEV that could achieve 80 mpg. I thought that structuring the program around diesel was a big mistake and that allowing the three domestic automakers more freedom in choosing their fuels and technologies would have been more productive. GM was interested in using our battery expertise to help in the development of their vehicle. AeroVironment was the primary subcontractor to GM in the PNGV program. We had a very small contract and were a subcontractor to AeroVironment.

The contractual paperwork for the PNGV work took some time to implement. I used a graduate student in this research, and we designed a horizontal, sealed lead acid battery for GM's HEV. When we were about a year into this development, GM called a meeting which included GM engineers, the people from AeroVironment, and other subcontractors including myself. The meeting was held in Denver, and I remember meeting many of the participants working on the project for the first time. The meeting took two days with the first day devoted to presenting the HEV's design and some of the problems they were encountering. The battery they had chosen to use in the vehicle was having difficulties. I was the last person to make a presentation that first day. I wasn't sure how interested people were in what I had to say but I did provide our design and discussed the benefits of the design.

The next morning, before anybody could start a presentation, the lead engineer for the project informed the project manager that he thought they should use my battery. I was shocked as I didn't even think people were paying much attention to what I said. After my presentation that first day, the engineers huddled together and held an impromptu meeting to discuss the design. Later they went to dinner together and continued their discussions. Besides solving the battery power and life problems, the horizontal plate design also provided better heat transfer and packaging for vehicle integration. After about 30 minutes of discussion on my battery, the GM manager for the PNGV program stated that changing battery designs was not going to happen as they did not have the schedule or funds to change the design. They were going to have to make their present design work.

The battery they were using had a cylindrical design, the cells looked like tall cans, and were popular with EV enthusiasts. Gates is a large automotive part manufacturer and was the battery company making these sealed, lead acid batteries. I worked with the battery engineer at AeroVironment to help him solve some of his problems. I suggested that he cut the battery cans in half to increase power performance and reduce the grid corrosion at the top terminals. Reducing the resistive losses in the batteries would also help with the heat transfer as not as much heat would be generated. Over the next few months, we talked quite a bit about solving his problems. He eventually was able to make the battery work and we stopped talking.

In any case, I believe I did help AeroVironment with the battery problems. I found out years later that Wally Rippel was also working for AeroVironment at the time all this was occurring. Although I was not a "big fan" of the PNGV program, I did think that GM developed the best PNGV vehicle out of all the companies involved in this program.

The other major battery development program for the horizontal plate battery was funded by ONR. The ONR program manager for our work also funded the joint tactical electric vehicle (JTEV) developed by AeroVironment, LLC. AeroVironment provided me with the ONR contact and we were able to generate a proposal in 2003 that led to a funded program for a hybrid electric Humvee in 2004. Although UI had the technical responsibility for this development, we worked with three companies previously involved with our research to develop the HEV Humvee. Concorde Battery Co. helped us design the horizontal plate battery so they could more easily fabricate the design and tooling needed for building the modules. AeroVironment tested individual modules for different driving cycles and tested a full vehicle battery pack for thermal and energy performance. UI purchased a 150 kw, ac drive system from AC propulsion, the company started by Alan Cocconi and Wally Rippel.

The horizontal plate battery and the AC drive were the powertrain components for the ONR HEV Humvee. One of my graduate students made a YouTube video of the Humvee driving around town and showed the vehicle "peeling rubber." The vehicle turned "a lot of heads" both in town and on the highway. It is interesting to note that the powertrain was initially designed for a very efficient electric vehicle but was finally used in a military HEV Humvee.

When Concorde Battery Company tested our modules for maximum power, they found that their maximum power tester did not have sufficient current to do the test. This was the only battery they had ever tested where this was the case. They had tested much larger batteries and one of the owners went out to the lab to witness the test. After the test, the owner touched the terminal and found it cool to the touch. He was impressed.

Dr. John Canning, who was a graduate student of mine and received his Ph.D. in Mechanical Engineering from UI, worked on many of my projects, including this one. He decided to do some life cycling on our horizontal plate battery where he would discharge a battery 100% on every cycle. After 240 of these cycles, his capacity had only dropped about 10%. When he got busy with other projects, he stopped the module cycling. This anecdotal information was encouraging.

With this history, I included pictures of the 12v horizontal plate design (see Fig.3), prototype modules (see Fig. 4), and the ONR HEV Humvee (see Fig. 5). ONR was very supportive of allowing UI to keep the HEV Humvee as a testbed to continue our research. Unfortunately, when they found the paperwork on the Humvee, they learned that the army owned the vehicle, not ONR. Because of the Iraq war, the Secretary of Defense had issued a general order that all Humvees would be repaired for possible use in Iraq. We found this out with about two months left in the program, so we had to strip everything out of the Humvee and ship it across the country to an Army facility in Pennsylvania. We had to literally winch the Humvee onto the carrier transporting it back East. Everybody at ONR and the Army that I talked to wanted to leave the vehicle at UI, but because of the general order they could not.

As we were finishing our HEV Humvee program, the Li-ion battery was gaining favor for EVs although these batteries were still prohibitively expensive. The military was interested in using Li-ion batteries in a military Humvee. I understood that a proposal for using Li-ion batteries in these vehicles had been submitted but I am not sure if this proposal was funded. In the last few years, the military has started to fast track the development of Li-ion for military vehicles including a hybrid electric Bradley Fighting Vehicle.

My present research interests at the University of Idaho includes fundamental work on the UI conductive carbon used on ceramic additives and investigating the horizontal plate battery as a low-cost energy storage system. These lead acid batteries could provide a low-cost solution for energy storage and one application for such a battery would be for high-rate charging of EV batteries. Our low-cost, high-power battery could be charged during low demand times, such as at night, and used during the day when the demand for EV charging is the greatest. In this manner, the power demanded during the day from the utility companies could be controlled while the high-rate, EV charging occurs. The batteries would be stored in a facility where EVs could be driven to the battery where the charging occurs. This approach could also be used for charging fleets of vehicles during the day when routes are determined, and charging can be scheduled.

The UI conductive carbon used on ceramic additives could dramatically improve energy performance for lead acid batteries. This conductive material that is stable at high potential could be a technical breakthrough and allow active material utilization higher than the present limit of about 50%, due to the low conductivity of the discharged material. Theoretically, the specific energy of these batteries could be increased from 30-40 w-hr./kg to 60-70 w-hr./kg by increasing the utilization of the active material. This improvement in active material conductivity is also expected to improve power performance and life. These low-cost batteries might be used in fleet vehicles where the range is limited and the opportunity for quick charging available.

GM and the EV Revolution

ZIERLER: How closely did you follow all of the exciting developments with the EV1?

EDWARDS: When I left JPL, after the ElectroSpirit proposal had not been funded, I thought that was the end of the EV development we were attempting. I was therefore very excited about the Impact electric vehicle and how well it was received by the public, January 1990. I was as excited when I found out from Wally that GM was going to make a production version of the Impact in April of 1990, the EV-1. He said this decision was made after a board member drove the Impact EV at GM's track and pronounced that "GM had to make this vehicle." The first EV1 vehicles were produced in 1996 and leased to the general public.

I think that GM did a good job on the EV1, although six years to develop a production vehicle from the prototype seems like a long time. EV1 did many good things for EVs. People used the vehicle primarily to commute to work for which it was well suited. The vehicle had great acceleration and handled well. The person leasing the vehicle did not have to stop at a gas station as they could charge at night. The electric cost for driving the vehicle was low. The vehicle was very dependable and required very little maintenance. We had tried to emphasize these operating features to promote EVs, but only after people experienced these features directly did they start to truly appreciate the value of EVs. Again, I was excited about all these developments.

Although the EV1 did achieve many goals and was very successful, I believe that using a trailer with a generator to provide for cross country travel would have eliminated the criticism that EVs could not travel very far. In the ElectroSpirit proposal, this trailer was referred to as "second wind." Having this cross-country capability would have been reassuring to many people who may never have used the feature. It would have answered the question "what can you do when you want to go a long distance." From my GM contacts, my understanding is that this feature was not included because it was an afterthought and the EV1 unibody construction made it difficult to find a good attachment point for the trailer.

ZIERLER: What are your thoughts on GM and their recall of the EV1?

EDWARDS: The criticism that I have for GM is that they didn't work very hard on EV technology to make it better for consumers. They had one of the most innovative companies in the world, AeroVironment, that they could tap into, and I don't think that they did. I realize that they were concerned about the California Air Resources Board (CARB) mandating EVs in their fleet requirements, so I understand some of their reluctance to advance this technology. I think that this was all unfortunate and should be a case study in how not to manage emerging technologies.

People were not happy when GM made the decision to stop leasing the EV-1 and began collecting and crushing these vehicles. I think the documentary "Who Killed the Electric Car" documented the feelings people had concerning GM's actions. Many people who were involved with the development of the Impact vehicle and who were Caltech graduates appeared in this documentary. GM had started producing the EV1s in 1996 and appeared happy with the good publicity although I don't think they believed they could make money selling them. GM was also afraid that making them better for consumers would cause demands by regulators to impose fleet requirements on selling them. Nobody appeared interested in making the EV better for consumers so they could sell more of them. Perhaps, if CARB had allowed lower mileage vehicles to be sold for every EV sold, then GM might have tried harder to make a better EV.

The documentary did discuss the problems that EVs had and which of these problems may have caused the demise of the EV1s. I knew many people who appeared in the documentary including Wally Rippel, Paul MacCready, Alan Cocconi, and Alec Brooks. Wally had contacted me about possibly appearing in the film although I was never asked. My contribution would have occurred prior to the history the film producers were interested in exploring.

The narrative of the movie suggests that the impetus for EV development began with the SunRaycer (1987) vehicle in the Australian solar car race. Wally Rippel believes that the ElectroSpirit proposal was the beginning of the EV development (1985) while I thought that our inability to fund the proposal was the end to our EV development. We had spent about eight years identifying and developing the critical EV technology prior to the ElectroSpirit proposal that was later used in the Impact EV. This technology included a high performance sealed, lead acid battery and an AC drive, along with a very efficient vehicle that AeroVironment could provide, and this technology was the basis for the ElectroSpirit proposal.

Of course, the simplest answer to the documentary's question was GM killed the EV. I believe GM's motive was to try to minimize the visibility of EVs and to fight CARB about implementing their EV mandate. I understand why GM did it and blame CARB as much as GM for this decision. The role that CARB played in the demise of the EV1 should be critically evaluated. A different strategy where CARB had a less adversarial and more collaborative relationship with GM may have yielded better results. Also, the competency of CARB should be assessed as their support for a hydrogen vehicle based on fuel cells was not warranted. I believe Alec Brooks fought this decision with some success.

ZIERLER: Did you ever get to drive the Impact or the EV1?

EDWARDS: I never had the opportunity to drive or even see the Impact in person. I saw pictures of it at the LA car show and I thought it was striking and beautiful. I never thought it would ever happen. I credit Paul MacCready for this small miracle. Without his dedication and perseverance, I don't believe the Impact would have ever happened.

I did have the opportunity to see an EV1 in person and even sat in it, but I didn't have the opportunity to drive one. UI was participating at an HEV competition in Phoenix where an EV1 was being displayed. Our students were involved in both HEV and EV competitions and this competition occurred around the time that the EV1 was starting to be produced. The organizers of the competition allowed me to sit in the vehicle but did not let anybody drive it. I remember that the batteries were in a tunnel which I did not particularly like because I felt somewhat constrained. Somewhere, I have a picture of the EV1and our UI HEV sitting head-to-head under the banner for the competition. I was not able to find that picture for this oral history.

ZIERLER: Did you think GM did a good job?

EDWARDS: Yeah, I think they did. It worked. I would've used our horizontal plate battery but obviously that was not necessary. They got a production vehicle out there. Once they got it out there, people started saying, "Hey, I kind of like this." I think it became a reality at that point, even though they didn't sell them, they just leased them. They crushed them. I think GM has always regretted that decision, but they were the first to make a viable production EV which was a huge accomplishment.

ZIERLER: What about the developments for AC Propulsion? Were you following what Alan Cocconi and Wally Rippel were doing at that point?

EDWARDS: Yeah. When I was up here, probably about my fifth year, '91, some students approached me about doing this student competition with electric and hybrid electric vehicles. I realized how much money and resources would be required, and I didn't think U of I could do it, so I kind of discouraged getting involved in it. But then, I had another group of students from WSU (Washington State University) approach me about doing this competition. I thought, "Between WSU and U of I, maybe we can do it." And WSU's only about 10 miles away from U of I.

We had the only joint school-sponsored hybrid electric vehicle team. And I must've been talking to Wally about this, he said, "Dean, you should buy one of these AC drives for your car." It was, like, $50,000. I said, "I don't think I can handle that." But then, we approached Schweitzer Engineering Laboratories, which makes a lot of equipment for the utility industry, about funding us. Ed Schweitzer was there, and we made a presentation, and he gave us the money to buy that drive.

My understanding is that Alan Cocconi and Wally made five drive systems, and they sold one to GM, some of the larger car companies, but they sold one to U of I for $50,000. We got one of the original drives. The UI HEV (1994) had the only AC drive in the competition (see Fig. 6). We could peel rubber. [Laugh] We actually broke an axle in that first competition because we had so much torque. We had to make some modifications to adjust for that failure. That hurt us, but we still ended up coming in third in that competition, even with a busted axle. It was fun. I was involved with a number of these student competitions.

We bought other systems from AC Propulsion. When we did the HEV Humvee, we bought their 150-kw drive system. We also bought their battery management system. That was with ONR (Office of Naval Research) money. We still have one of their 150 kw drive systems, battery management system, and bidirectional charging system. With funding from Avista Power Co. we used their bi-directional charging capability for measuring noise on the utility grid. The systems all work, and we have them installed in an HEV chassis testbed (see Fig. 7) with a pack of the horizontal plate batteries from the ONR project installed under the floor.

AC Propulsion was the company doing the innovations for electric vehicles. Alan Cocconi demonstrated a range extender when he traveled from Los Angeles to Washington D.C. and back again in 1995. The t-Zero produced by AC Propulsion in 1997 had a range extender designed for that vehicle called the "long ranger." With these range extenders, the EVs could easily be changed over to a vehicle having a very long range. Alan and Wally also pioneered EV bi-directional charging which I believe will be critical for load leveling and long-term energy management. They also have been involved with fast charging, which is important for many different EVs including fleet vehicles. I mainly interacted with Alan and Paul Carosa. They were all very competent and innovative engineers. I believe that AC Propulsion was the innovative center for EV research starting from when they were founded until the beginning of Tesla.

ZIERLER: What about the development of lithium-ion batteries? Were you involved in that at all?

EDWARDS: We talked about that. I was always skeptical because they're so expensive. And they still are. It wasn't that you couldn't get the performance out of it, but, because of the cost, I just didn't see that as being a way of penetrating the market. The major reason for pursuing lead acid batteries is their low cost and ability to be recycled. I was trying to improve the lead acid battery to increase both energy and power performance while keeping the cost low.

But yeah, Wally talked about that. I understand that Alan Cocconi wanted a 300-mile-range electric car and didn't care what it cost. He decided that the Li-ion battery was the only way to get it. And he did it, that was it. It was a high-performance thing.

People will pay for performance for things they are interested in having. You have these expensive golf clubs, carbon filament, all this technology that is expensive but what people really want and will pay for. People wanted this vehicle that could go 300 miles, but it could also really leave anybody in the dust, and I think that was a good strategy. People wanted to buy Teslas both for the range and power performance and that is what created the EV market.

At about the time that GM was recalling and crushing the EV1s (2002-2003), Alan Cocconi was producing the first EV with a Li-ion battery. AC propulsion showed a Li-ion battery EV in September 2003. I was excited about this 300-mile range EV although the high cost of the vehicle made me question what the size of the market would be for such a vehicle. I always assumed that the crushing of the EV1s provided Alan with the motivation to design and build this vehicle. However, these two events may only be coincidental as the Li-ion batteries were starting to be used in power tools, laptop computers, and radio-controlled hobby vehicles at about this time. The timing of these batteries becoming available and the discontinuation of the EV1s may only have been coincidental.

The development of this expensive electric vehicle that had long range was interesting but not necessarily important except that the Li-ion batteries also had long life. Wally told me that Alan had looked at the life of these batteries and found that the amount of energy used over a battery's lifetime was large so that the battery lifetime cost was almost competitive with other batteries. Although the vehicles had a high initial cost, the fuel (i.e., electricity) cost, the maintenance cost, and the battery replacement cost based on energy usage showed potential for making the vehicle lifetime costs competitive with conventional vehicles. This insight was important and resulted in another innovation by AC Propulsion, the EV Li-ion battery.

ZIERLER: Beyond the cost and the performance, did you recognize early on that lithium-ion would be a solution for so-called range anxiety in EVs?

EDWARDS: We called that "island fever." You're on this island and you can't get away. It was the problem of having EVs that wouldn't go so far. With the trailer, that was our way to address it. Basically, you drive it as an EV in town, during commuting. If you want to go cross-country, and you don't have another vehicle, then you either own a trailer or rent one, hook it up, and use the small IC engine with generator to provide you cross-country capability. That was our approach. It's hard to beat the internal combustion engine for energy density. That was our strategy for getting around the range limitation.

The Tesla is more of a straight substitute for the conventional vehicle as it has a long EV range, consistent with conventional vehicles. However, the problem is now that charging a Tesla takes a long time whereas filling a gas tank can be done very quickly. Long range EVs are still not a straight substitute for a conventional vehicle. The EV range problem has now been replaced by the charge time limitation. The small trailer with motor generator would solve both the range and charge time problems of EVs.

ZIERLER: Were you following the developments when Martin Eberhard, and later, Elon Musk, took an interest in AC Propulsion?

EDWARDS: Not really. Other than I talked to Wally, and I knew he had started working for them. I had talked to him about that, and Alan had demonstrated the 300-mile-range vehicle. We talked about this stuff in general and using lithium-ion batteries in an expensive sports car. I thought all that made sense and that a market existed for the Tesla EV. I just didn't think it was very large but I thought the Tesla EV was amazing.

Wally always thought that the large-scale battery consumer market would eventually create a better battery for EVs. You have this market out there for high energy batteries, the computer market. That's going to help support the development of these types of batteries. It was on our minds, but I viewed the big problem as being cost. There are technical problems, but I figured you could solve those problems. Eventually, it would come down to being too expensive for most people. That's why I tried to develop a less-expensive technology that would be able to provide enough performance for EVs to "get the job done."

ZIERLER: Which was a different kind of battery?

EDWARDS: The horizontal plate, sealed lead acid battery had the power performance and cycle life needed for a successful EV as demonstrated by the Impact and EV1. Because the conductivity of both the positive and negative plate changes as the active material is converted to lead sulfate, which is non-conductive, the maximum utilization, even at low rates, is about 50-55% utilization. At high rates of discharge, the utilization can be limited by diffusion and be even lower. We started looking at different additives, both conductive and porous, to try to overcome these physical limitations and improve the utilization of the lead acid battery. This was the different kind of battery we wanted to develop.

The UI carbon material that Dr. Cheng in the Chemistry Department discovered by accident could be very important for improving the performance and life of our battery. The material can withstand the high-voltage environment of a lead acid battery and provide a conductive structure. The ceramic fiber when coated with this carbon could possibly achieve 70% utilization with less than 20% additive volume, as predicted from computer models. The porous hollow glass microsphere (PHGM) can also be coated with this carbon material. This conductive additive could provide structure to store electrolyte inside the plate to improve diffusion without the attendant drop in performance when non-conductive additives are used.

I hoped to try to develop a sealed lead acid battery to get, instead of 25- or 35-watt hours per kilogram, 40 to 60 watt hours per kilogram at high rates. That would put it within the range of the original lithium-ion battery systems. Then, you could increase the range of the EV with the same size of battery while keeping the cost low. That was a lot of my research, doing modeling and looking at these additives. The modeling would estimate what kind of utilization and power performance you'd get out of that system.

The Legacy of Sealed Lead Acid Batteries

ZIERLER: What happened to the sealed lead acid batteries? Were they adopted in any applications?

EDWARDS: They're used in many applications and may be the most common type of storage battery. Their biggest use is probably in uninterruptible power supplies, but they are used in many automotive and telecommunications systems. They are used in some deep cycled applications, but this is not a large market for them. Of course, we were interested in developing a deep cycled, sealed lead acid battery that had both good power and energy performance with a long life. We believe that the use of conductive additives would allow us to create the additives needed for this battery. Exide was interested in the porous hollow glass microspheres (PHGMs) but we found that these additives needed to be conductive to achieve the full benefit of improving diffusion. We successfully coated the PHGM with the UI carbon but did not have enough quantities of these coated PHGMs to perform cell tests.

We did work with the UI conductive carbon and believe we can eventually achieve utilization of the active material greater than 50%, even at high rates of discharge. There's still quite a bit of energy you can get out of a lead acid battery if you can figure out how to do it, but that requires getting money. We presently have some proposals to try to get some money to demonstrate the increase in utilization. Finding this money is not easy. A lot of applications with sealed lead acid batteries are well-defined, so they're not looking at high performance. They have a battery that has dimensions defined and a certain amount of power, so it's hard to improve anything for these specific applications. Sponsors are interested in batteries for meeting these applications and not research.

I think the areas that would make the most sense for the high power, horizontal plate battery would be for storing energy to be used for fast charging of EVs. These batteries could be used for fleet applications where the low cost, horizontal plate battery would be charged over an extended time and then quickly discharged to fast charge vehicles in a fleet. We are working on a proposal for doing some of this work. But again, sponsors want a battery, and do not want to do research to develop technology that may enhance battery performance at some time in the future. We want to develop and demonstrate a battery that can be used for fast charging fleet EVs.

ZIERLER: So I understand, the main attraction here is simply cost, that sealed lead acid batteries are so much less expensive, and therefore so much more attainable in a variety of applications than lithium ion?

EDWARDS: Yes, mainly cost. This is particularly true for stationary energy storage applications where weight and energy density are not as important as in electric vehicle applications. The sealed lead acid battery has a low initial cost and low maintenance costs due to its sealed design. The horizontal plate battery has excellent cycle life so that the total lifetime cost of the battery should be low. The high-power battery should also be able to compensate for sudden demands from the grid. These same characteristics would make this battery a good candidate for fast charging of electric vehicles, which is the application we are presently pursuing.

The horizontal plate battery, particularly one with improved energy density, could provide a 100-mile-range electric vehicle with a much less expensive battery, as shown by the Impact and EV1 vehicles. The vehicle could be used primarily for commuting and could be charged at work or wherever you stop. The charging may only be partial, but it would add to the EV range. For cross country travel, a trailer with a motor generator could be used. But the main thing is reducing the price of EVs so most people can afford them and making EVs as easy to use as gas powered cars.

The fact that you can get $10,000 from the government to buy an EV does not help to develop lower cost EVs. It helps people buy an expensive vehicle but at a lower price. The person who's going to use this for utilitarian reasons doesn't want to spend that kind of money. The idea is that a lower priced EV would have a greater chance to penetrate the market. I still believe this is an important technology to develop although it may not be high tech.

Back to top

Idaho and Robotics

ZIERLER: What about your robotics research, were you able to pursue that at UI?

EDWARDS: I did but it was not easy. My first research project in robotics was an autonomous forest vehicle. These robots were initially meant to provide the same capabilities as horses which historically had been used for "skidding" logs to a landing where they were loaded onto trucks. These forest robots were envisioned to be small, low cost, and autonomous. They could follow skid trails by themselves, so no operator was required for driving which reduced labor costs. Both the capitalization and labor costs were low for these vehicles.

I got the idea for these robots when I bought a teleoperated toy bulldozer for my four-year-old son. The thought occurred to me that if you made the toy larger, you could perform useful work. My interest in forest robots was a result of my dad's experiences. He was a UI Forestry graduate and traveled from Massachusetts in 1935 to attend the UI College of Forestry (COF). During his first summer, he worked in a logging camp near Bovill, ID which used horses for logging. For small logging operations, horse logging can still be competitive with more conventional types of logging. The robots were to be small, low-cost vehicles that were autonomous for most of their operation.

When I started the forest robot research, I had no students and no funding. Through some personal contacts, I was able to secure $10k of funding from Potlatch Forest Corp. and Bennett Lumber Co. With this money, I was able to sponsor a forest robot project in my robotics class. The funds were used to purchase a small, manual operated forest skidder, called the Iron Horse, which was made in Sweden by Husqvarna. The students in the class were clever and were able to modify the actuators so a hobby grade radio controller made by Futaba could be used to remotely operate the vehicle. Although small, the radio controlled (RC) skidder was capable of skidding large logs (see Fig. 8).

The next iteration on this vehicle was to provide for autonomous operation so the vehicle could follow a skid trail. We developed a hierarchical fuzzy logic control system that used encoders on the track shafts, an array of six ultrasonic sensors, and a GPS receiver. Initially, computer simulations were used to evaluate and optimize this control system. An instrumentation platform, which was mounted on a baby carriage and included all the control components minus the actuators, was also used to evaluate the controller. We called this platform "fuzzy baby" which was short for fuzzy logic baby stroller (see Fig. 9) and would use this platform in all our robotic developments, even autonomous underwater vehicles (AUVs).

We modified the small RC skidder by replacing the clutch-brake skid steer with a hydraulic drive. Although we did try to use GPS in our controller, the forest canopy and terrain usually prevented the signal from being received and was not very accurate when a signal was received. However, we were able to implement our fuzzy logic hierarchical controller on the vehicle where an array of six ultrasonic sensors were used to "see" the trail and had encoders on the tracks to monitor distance. The control system used a control run where the operator would use radio control to drive the robot down a trail. The data from this run would be compared with the data being sensed when under autonomous control. This approach was successful as we could combine different strategies for following the trail. The fuzzy logic controller worked very well for our operations.

With this success, we approached USDA (United States Department of Agriculture) about funding this technology. Although they usually did not fund equipment development, they did provide us with a small amount of funds to continue our forest robot research. With these funds, we bought the smallest two-tracked vehicle we could find, an ASV-30 (see Fig. 10). Because the vehicle used manual control, I made the decision to keep this capability so the manual control could be compared with radio control as well as autonomous control.

Interestingly, the operators preferred to drive the vehicle by radio control rather than manually and would walk along with the vehicle rather than ride in it. The small vehicle gave a rough ride and provided the operator with limited vision. They would choose to remain outside the vehicle where they could see better and avoid the rough ride. The vehicle is still used by the College of Natural Resources in different studies.

The USDA research was very successful. We implemented our hierarchal fuzzy logic control system on a forest vehicle that could skid logs and operate under radio control and autonomous control. Our vision system of six ultrasonic sensors was mounted on the ASV-30 vehicle in an array (see Fig. 10) and provided distances to nearest obstacles on and near the path for autonomous control. We demonstrated the autonomous control on skid trails up to 100 yards. The fuzzy logic control system could combine different control strategies and could be trained and would learn from runs down the trail. Although additional research was required before this technology could be implemented, the funded research had provided proof of concept.

USDA did agree that our work demonstrated feasibility and they encouraged us to seek funding through a small business initiation research (SBIR) grant. In order to apply for these SBIR grants, I founded a company, Forest Robots, LLC. Through this company, I was able to apply and was awarded both a Phase I and II SBIR grant. In Phase I, we implemented a more powerful antennae for our radio control so we could maintain control of our vehicle out to 1000 ft even through the forest canopy. In Phase II, we developed a vehicle called Forest Crawler (see Fig. 11), and a commercialization plan. Unfortunately, the forest robot was not very close to commercialization, and this was true even at the end of Phase II. No more funds were available to us after the Phase II SBIR.

The purpose of the small, semi-autonomous equipment being developed at Forest Robots was to reduce the high fuel loadings in our forests. This is a huge problem that our SBIR proposals addressed. The present fuel loadings are responsible for the devastating fires that occur every summer resulting in loss of property and sometimes lives, bad air quality, and the release of tremendous amounts of CO2 into the atmosphere. These small forest robots have the potential to reduce forest fuels in the most cost-effective manner thereby allowing for more remediation to occur. Various attachments could be used on these robotic vehicles to deal with brush and small trees. The wood waste could then be chipped or pelletized and used for energy. In this way, useful energy could be gained from all the CO2 that wildfires produce.

Just about every summer when the skies are smoky and the weather is hot, I think what a shame we didn't see this research through to a better conclusion. Forest fires burn millions of acres every year which results in millions of metric tons of carbon dioxide being released into the atmosphere. In 2020, California had a record-breaking fire season which resulted in the release of an estimated 127 million metric tons of carbon dioxide. This release represented almost twice the reduction in carbon dioxide achieved by the state over the previous seventeen years. Ideally, harvesting this material and using it for energy would allow a tremendous amount of energy to be generated with a carbon neutral source. Although given the opportunity, I would have done things differently in Phase II, but I'm not sure the results would have changed.

During the USDA project, we worked with many different people including professors and students in engineering and the UI Forest Products Department. The College of Natural Resources had an experimental forest and a student work team that would log and conduct tests during the summer. They were extremely helpful and cooperative in helping us develop our vehicles. We also interacted with the forest industries, as previously mentioned, and with the Forest Service. We worked with the Rocky Mountain Research Station which is a center of the Forest Service located in Moscow, Idaho. We also worked with people in the Forest Service back in D.C. who were very supportive of our research and what we were trying to accomplish.

We did receive state funding to test two vision systems for our forest robot. One system used a single, color camera for identifying a forest trail. In another system, we used a 3D vision system to guide our vehicle down a forest trail using a fuzzy logic, hierarchical control system. This vehicle could repetitively drive the robot down a 400-yard trail and stay within a couple of feet of the center of the trail during the entire run. These results showed that small forest vehicles could operate autonomously in the forest to do fuel reduction remediation for fire suppression.

Another autonomous vehicle program in which we participated was sponsored by DARPA (Defense Advanced Research Project Agency). The program was named LAGR (Learning Applied to Ground Robots) and each team that participated received the same robot (see Fig. 12), so that only the software and control strategy was what differentiated the teams. The teams included some well-known participants including JPL, Stanford, SRI International, Georgia Tech, etc.

The Office of Naval Research (ONR) sponsored our participation in this program. The program was structured to not be competitive so that software could be shared, and teams could benefit from other people's success. We were one of two teams to finish in one of the last runs we made on a difficult course. Our algorithms were developed for forest navigation and used forest terrain and natural vegetation for training. In this run, the vehicles needed to traverse through some light brush which stopped many of the vehicles but was not a problem for our vehicle, which was trained on similar terrain.

The largest research program that I directed was sponsored by ONR and involved a fleet of autonomous underwater vehicles (AUVs). This fleet of AUVs (see Fig.13), could communicate and cooperate with each other to perform missions. One of these missions was to sweep an area for mines. The AUVs would work together in a formation to search for mines and, when a possible mine was found, would deploy a vehicle to inspect the mine-like-object and verify its existence and location. The mines identified in this manner could then be neutralized in future operations when necessary.

Another mission the fleet of AUVs performed was to make magnetic and electric field measurements on ships. They would move in a close formation under a ship and make measurements which could then be combined to assess the overall signature of the ship. Both these missions, sweeping for mines and ship measurements, required formation flying, deployment, retrieval, vehicle replacement, formation reconfiguration, and communication between the different vehicles as well as with the operator. The AUVs had the language and the associated logic for these missions. We used computer simulations to evaluate how the AUVs would behave on specific missions and different scenarios. We also performed tests with multiple AUVs with a more limited scenario to test how the vehicles would communicate and behave in the water. These tests were performed at the Bayview Acoustic Research Detachment located on Lake Pend Oreille (see Fig. 13). This facility is only about a two-hour drive from UI and was where we did most of our testing.

A type of artificial intelligence, which we called Language Centered Intelligence (LCI), evolved from our research on multiple autonomous underwater vehicles (AUVs). A language was invented to implement these AUV missions that we called AUVish (like English, but for AUVs). This language and associated logic support the communication and cooperation needed for conducting multiple AUV missions. While doing this work, we realized that different future scenarios could be evaluated by using AUVish and the logic associated with it.

For instance, a series of communications and decisions are triggered when a hypothetical mine-like object is discovered and reported. The response of the AUVs can be determined from this hypothetical situation where messages and commands are communicated. This information can be used to evaluate the response and to optimize system assets for hypothetical situations. Of course, this hypothetical situation would need to be identified and separated from the real communications.

The LCI approach to developing and evaluating these scenarios is similar to the way humans' process information and think. Try formulating a thought without using words or language – it's very difficult. Humans think and reason using the same words and language that are used for communication. Unlike most computerized systems, in which communications is a layer independent of reasoning, human intelligence and language are closely, almost seamlessly, connected. Clearly, we are the most intelligent animal on this planet and also the one having the most developed language. While visualization is used in some thinking processes, even these processes make use of words to help define or explain the reasoning process.

Although this approach could be used with natural language, we limited our investigations to the artificial language (AUVish). LCI is the reasoning techniques associated with a language, in our case AUVish, and its logic. LCI is the integration of language and logic into an efficient and effective way for AUVs to autonomously conduct mission planning, reason about future scenarios, respond accordingly, and communicate their reasoning to humans-in-the-loop. Because of the inherent limitations of the AUVs and the constrained domains of the fleet missions, AUVish is necessarily designed to be much simpler than natural languages such as English but it shares many of the same features.

Surveying the EV Landscape

ZIERLER: I wonder what your overall take is on the status of EVs right now, both in terms of the technology and their adoption among the consumer marketplaces.

EDWARDS: Right now, the Tesla EVs are well engineered and have remarkable performance. I think the AC drive is what separates electric vehicles from gas-powered cars in a good way. EVs have great acceleration and handling performance. The limited range problem of previous EVs is now being replaced by the long charge time problems for the high-capacity EV batteries. Unfortunately, the overall price for an EV is quite a bit higher than a gas-powered vehicle. Most of this high cost is associated with the lithium-ion batteries that use a lot of expensive metals only found in certain locations. The price of EVs must be reduced to increase the market penetration. The cost, range, and charge time problems are all related and the market should provide different solutions for different customer needs.

ZIERLER: Do you think adoption levels are going to get to a point where this meaningfully impacts carbon emission reductions?

EDWARDS: My view is that EVs should be able to help in meeting carbon emission goals. When used with nuclear energy, EVs could dramatically reduce carbon and, with bidirectional charging, could be valuable in performing load leveling for nuclear power plants and maybe renewable energy sources. If we are going to cut carbon emissions, I believe we need more nuclear energy. A positive net energy flow has been demonstrated from fusion experiments and I believe we will eventually transition to fusion nuclear power plants. EVs work well with these nuclear energy scenarios.

I don't believe we have the time to try to make renewables provide all our energy. We are finding they have environmental problems just like with dams and hydropower. We need a carbon free energy source as soon as possible and I believe nuclear energy is our only choice. Eventually we will be able to transition to fusion nuclear power over the next twenty years or so. A mix of 40% nuclear, 30% renewable, and 30% natural gas would minimize the carbon dioxide generated for electricity. In this scenario 70% of the generated electricity would be carbon free and the natural gas power plants can attain very high efficiencies with a minimum amount of carbon. Our present mix is 60% coal and natural gas, 20% nuclear, and 20% renewables. I believe 40% nuclear and 30% renewables is a realistic goal and one which would reduce our carbon dioxide generated to less than half what we have today. I do cringe at my nuclear energy suggestion because of problems associated with waste products and safety, but I now believe this is the best path forward to dramatically reduce the carbon dioxide we are generating.

I know they're trying to get rid of internal combustion vehicles and only sell electric vehicles, but I don't agree with that approach. They're getting rid of vehicles when what they want to get rid of is carbon. You want to tax the carbon going into the atmosphere. If you do that, and people become sensitive to that cost, they'll react to it. They may carpool or telecommute to work, or use public transportation, etc. There are lots of different ways to reduce carbon dioxide, and you'd like to do it as cheaply as possible. That's my feeling. As you increase the price on carbon, electric vehicles look more and more attractive. Right now, I think they're trying to get electric vehicles out there with subsidies, thinking that's going to reduce the cost. I don't think that's necessarily true. We need to work on reducing the cost of EVs and solving the charging problems so people will buy more of them.

ZIERLER: What about simply the cost of EVs? They're certainly not attainable for everybody. Is that part of the equation as well? Do you think prices will come down?

EDWARDS: I don't think so. At least for the near term, I think they're going to be high. The battery's going to be costly. The AC drive and the maintenance costs on these vehicles are fairly low. The energy costs are low, sometimes because they're not paying gasoline tax. But I don't ever see the price of the electric vehicles coming down to where they could compete with internal combustion vehicles as far as initial cost is concerned. However, the lifetime cost of EVs may be competitive with gasoline cars and is something that should be considered when comparing the cost of EVs with gas vehicles. For specialized EVs, the initial cost could be substantially reduced. If I'm commuting 50 miles or less every day, I would want an electric vehicle. I may not be able to buy an expensive one, but if you don't have to drive 300 miles in your vehicle, then you don't have to buy a big battery. Maybe you can buy a smaller battery, and that would work better.

In the original Electro-Spirit proposal, we had a limited-range electric vehicle. This EV had over a hundred-mile range and could be used for 80-90% of daily commuting distances. We proposed having a trailer with a motor generator set that you could rent and charge the battery as you go cross-country. In fact, Alan Cocconi did that with his sealed lead acid vehicle. He had a trailer. There are "lots of ways to skin a cat." I don't think it behooves anybody to force one solution. I think the best thing is everyone finding their own solution as cheaply as they can.

ZIERLER: Let's go back in history. Before you were a graduate student at Caltech, when you were at Illinois Institute of Technology, tell me about what you were looking for in a graduate school and how Caltech got on your radar.

EDWARDS: I wanted a good technical school, and I was interested in controls. I applied to a number of schools and was accepted at all of them, although at that time, I was only interested in a master's, which was a big mistake. You don't get money if you just apply for a master's, you have to apply to a PhD program. It came down to Stanford and Caltech. And Stanford probably had a better controls program, they had a larger faculty involved in that area, quite a bit of money.

From that point of view, it may have been a better school for what I wanted to do, but I liked Caltech an awful lot for many reasons. Its tradition, the fact it's a small school. The ability to interact with other students and faculty from other disciplines. I decided to go with Caltech. The first year I was there, I paid tuition. I was just going to get a master's. I liked it so much that I decided I was going to stay and pursue a PhD. When I got into the PhD program, I started getting financial support.

Control Systems at Caltech

ZIERLER: What year did you arrive at Caltech?

EDWARDS: 1972.

ZIERLER: Into what program?

EDWARDS: It was mechanical engineering. I was interested in control systems, system analysis, including nonlinear system analysis. Mainly, I was interested in becoming involved in robotics.

ZIERLER: Were you aware of the interest in EVs at Caltech at that point? Did you know about Wally Rippel and the Great Electric Car Race of '68?

EDWARDS: No, I didn't. I got involved, to some extent, with Paul MacCready before he set the record for human flight. I became interested in hang-gliding through some people at Caltech, so our paths kind of crossed at that point. I knew his sons through some associations I had at Caltech with people who were interested in hang-gliding at the time. To me, Caltech was famous for aerodynamics and the space program, which included JPL. I associated physics and space with Caltech, not electric vehicles.

But I was interested in energy. In the 70s, they had one energy crisis after another. I just was not aware of what Caltech was doing in this area.

ZIERLER: Do you remember smog, and was that a motivating factor for you in thinking about clean energy?

EDWARDS: Yeah, when I got to Caltech, I was surprised how smoggy it was at times. Arriving late in summer, August, it's not the best time for smog. I played sports, and I was involved with some intramural stuff, and it was kind of tough with the smog. It actually hurt my eyes. That was certainly on my radar.

ZIERLER: Tell me about developing your thesis topic. What were you interested in as a grad student at Caltech?

EDWARDS: I was interested in control systems. My major professor, Tom Caughey, suggested that I analyze a group of control systems that use switching power semiconductors to efficiently control the energy flow for voltage regulators. He had been involved in a thesis defense and thought that time-domain analysis of these regulators was a better way of doing it. He also thought the time-domain methods could also be extended to do nonlinear analysis on these systems. We did a time-domain analysis on these switching regulators that was very successful. This is the same sort of power device you'd use, like in an inverter or DC-DC regulator. That's how I got involved with power electronics, through my thesis topic. Interestingly, Wally, Alan, and I were somewhat connected to the power electronics program at Caltech. A critical technology for efficiently handling electrical power.

ZIERLER: Tell me about lab work and instrumentation you were exposed to at Caltech.

EDWARDS: My thesis was mainly theoretical and mathematical. I used computer simulations and the available literature in my research so lab work and instrumentation were not critical for my thesis. However, Caltech was very helpful and accommodating to students working in labs and shops. In one of my first classes at Caltech, I became interested in hand gliders. Peter Lissaman, who worked for AeroVironment, was also an adjunct professor at Caltech and had developed this profile for an airfoil that gave high lift at low speed, so it was ideal for hang-gliding. I became interested in this special airfoil and decided to make a wooden airfoil with this profile, which I did. Unfortunately, I never had the opportunity to test the airfoil and it ended up in my closet.

The following year, a new roommate moved in with me who was involved in aeronautics. Someone in his class wanted to test an airfoil in the wind tunnel, and Caltech didn't have one. He told the other student, "I know where there's an airfoil." He got mine out, took it down there, and they ended up testing it in a wind tunnel, so it finally did get tested. The thing I remember about Caltech in that instance was being able to go into that shop and, with very little problem, get the wood, make the cuts, sand it down, and do everything needed to finish that airfoil. I'm not certain this activity was directly related to any of my classes but everybody in the shop was helpful. It was all a good experience.

ZIERLER: Who were some of the professors you were close with at Caltech?

EDWARDS: Tom Caughey was my major professor, so he was the closest one. But they had the applied mechanics group where he was a faculty member. I was up on the fourth floor where they had all these applied mechanics professors including Knowles and Sternberg. It was a very interesting place. The students joked that when these professors went for coffee, the IQ of the building dropped in half.

Tom was a brilliant individual and very personable but tough. When I read his oral history from Caltech, I discovered that he was the first student at Caltech in the Engineering Science program, which had been started by the famous rocket scientist Hsue-Shen Tsien. From his comments, he apparently was a graduate student of Tsien, who returned to China in 1955. Tom received his Ph.D. in 1954 just before Tsien left for China.

When I found out that I was going to Caltech, I bought a biography of the famous Caltech scientist Theodore von Karman. In the book, von Karman recounts how he was a student of Ludwig Prandtl and that Tsien was his student. During WWII, Prandtl worked for the Germans and von Karman and Tsien worked for the United States. Tsien subsequently left for China to work for the Chinese Communists. Tsien became well known in China for his work on their rockets and their atomic bomb. These three individuals were all brilliant scientists but served very different political masters. Tom was part of that Caltech history, which was very interesting to me.

I do remember interacting with many different faculty members. On Friday afternoons, I played basketball at the Caltech gym with graduate students and faculty, many associated with the Environmental Science program. Afterwards, we would go to the Athenaeum and drink some beers and then head out to dinner. This was all very social and fun, but I did get to know what research the students and faculty were doing. I played intramural sports, both baseball and basketball, and after games we would socialize and find out what people were doing. We had an intramural basketball team called the "Misfits," and Roger Knoll and a couple of his graduate students played on the team. Again, we would socialize and get to know the people and usually something about their research. In this instance, the research was economics, and they were analyzing sport teams and the associated economics of professional sports.

Because all the graduate classes were small, the students had a good opportunity to interact with the faculty teaching these classes. My first semester at Caltech, I had a thermodynamics course taught by Rolf Sabersky. I had read an article about the influence of carbon dioxide on the atmosphere and the greenhouse effect. I asked Professor Sabersky about this phenomenon, and he said, "it should be taken seriously." I did, and I may have been one of the few people around who was wondering why the atmosphere was not getting hotter. In the early 1990s, the temperature of the Earth started to get noticeably hotter, which verified global warming to me some 20 years after I started looking for it.

ZIERLER: Was your cohort competitive? Were people looking to best each other in their projects?

EDWARDS: No, I didn't get that at all. It was kind of interesting seeing some projects that were going on. One professor was investigating the dynamics of swimming fish. A lot of interesting research was being done. When you get to know the graduate students, you get to know the research they're doing and what areas they are investigating. I would say it was more interesting and fun than competitive.

ZIERLER: Tell me about some of the key conclusions of your thesis. What did you see as your contributions with this work?

EDWARDS: The use of time domain methods to analyze the switching regulators was new and provided new insight into these systems. You could get very good results using this analysis. I think at that time, people were having trouble doing the analysis in the frequency domain where the results could be hard to interpret. The time domain methods were more straightforward and would give you better results and insight into what was going on with these systems. We also extended these methods to analyze some nonlinear behavior of these systems.

ZIERLER: Who was on your thesis committee?

EDWARDS: Tom Caughey, my major professor. Dave Middlebrook, who was in electrical engineering and was involved with power electronics and switching regulators. I believe Professor Knowles and Sabersky were also on my committee. There was also a history professor, James Quirk, on my committee. Evidently, I had to have some credits in the humanities, which I had, but apparently not the right ones. Professor Quirk, who was an economics professor agreed to help me out by having me review a paper he had written about bison extinction. He did an economic analysis of bison and the hunting of bison. I read the paper and he questioned me about it. One of the results of his analysis was that after the introduction of the horse, the American Indian probably would've driven the buffalo to extinction. It depends on the parameters and other assumptions, but that result was a very likely outcome. Again, I found all this research to be very interesting.

ZIERLER: After you defended, was the opportunity at JPL all wrapped up for you, or you were on the market more generally?

EDWARDS: I had been at JPL the previous summer, so I had some connections there. I was also interested in other companies, including GM. However, Tom Caughey had advised me to not continue doing the same research as I did for my thesis. He thought it would limit me and cause people to question my ability to do my own research. I believe this advice was mainly meant for academic careers, but I did take it seriously.

At JPL, I did not interview where I had previously worked and was offered a job in another section. I received the job offer very quickly after interviewing and before I had interviewed other places. I accepted the JPL job offer although I would have liked to interview other places. Because I had previously arranged to interview and visit GM before I received the job offer from JPL, I did make that trip. Unfortunately, JPL didn't hire me as a PhD because I defended in late fall but didn't receive the PhD until that spring when I graduated. I worked in the spring at JPL, but not as a PhD even though I had fulfilled all the conditions. If I had been hired in as a PhD, I'd have started at a higher salary, which would've helped me. But I needed the money and the work, so I did not have a lot of choice.

ZIERLER: What group did you join at JPL? Did you have a good idea who you'd be working with?

EDWARDS: Yeah, I hired into Division 35 which does more mechanical engineering work. I worked in mission planning when I first started. I was involved with the space program and did some work with hydrogen peroxide rockets. After about a year, I moved over into the energy area, when I got the opportunity. DOE started funding work at JPL, and I started doing some EV testing for them. When I started doing this work, I moved to Division 34 into the power electronics group. The same group where I had previously worked. Eventually, I started doing work with batteries, drive systems, and other EV development efforts.

ZIERLER: On your work in power electronics, where was that applied in JPL? What missions was it used for, what research was it relevant to?

EDWARDS: Any time you have electronics, you have some sort of regulator to provide the correct voltage for the housekeeping power supply. The switching regulators, sometimes referred to as DC-DC converters, do this very efficiently by turning the power on and off very rapidly to minimize losses. On a spacecraft all the scientific payloads as well as the power system would use these switching power devices to minimize losses, which is very important when energy is at a premium. They are used extensively throughout a spacecraft.

The power used in electric vehicles is substantially greater than that used in spacecrafts. The same principles apply but the power semiconductors are much larger. That work was something we had to grow internally. Initially we did that with some internal funding. We did the same thing with the battery work, although we got the Electric Power Research Institute to start to fund that work. That was separate from DOE work and NASA.

Why Caltech is Central to EV History

ZIERLER: Now that we've worked right up to the present, for the last part of our talk, a few retrospective questions, and we'll end looking to the future. First, between Caltech and JPL, I've really come to appreciate how much of the development of EVs is a Pasadena story between the two. I wonder if you've ever reflected on why. Why have these institutions led the way in so many regards in this area? What does that say about Caltech?

EDWARDS: I think the people at Caltech are bright and knowledgeable. And I think those are the types of people willing to take on this challenge. Paul MacCready was interested in gliders and human-powered flight, but that had application to vehicles and electric vehicles. He got kind of pulled into that. He wasn't trying to develop an electric vehicle, but he had the technology, training, and insight to be able to make that electric vehicle very efficient. Paul MacCready, Wally Ripple, Alan Cocconi, Alec Brooks, and I were all Caltech graduates and interested in energy technology. I believe Caltech's philosophy of having a small community of very capable people that can interact is what happened with the Impact and EVs. We all had different backgrounds and interests but our ability to interact resulted in technology that I don't believe would have been developed anywhere else.

Caltech did have a power electronics program, although I don't know if that contributed directly to the success of the EV program. Power electronics was important to EV development, particularly the AC drive. The AC drive really helped to sell EVs. All of these circumstances kind of worked together. But it was really those people who were able to interact at Caltech, JPL, and AeroVironment that made this happen. Caltech admitted these people and JPL hired some of these people, but it was the way they interacted that was key to the whole thing. Eventually, that was what got it done.

ZIERLER: Well, the institutions are nothing if not made up of the people that who belong to them.

EDWARDS: Yeah. Just about all those people were pretty much required to get all this done, I think. But it's still just a handful of people. I worked with the Partnership for New Generation Vehicles. I think they spent, and I could be mistaken on this, something like half a billion dollars on these three car companies to come up with these designs for a hybrid vehicle that was based on a diesel engine and high gas mileage. I don't believe it ever went anywhere.

Yet, with a fraction of the cost, Alan Cocconi said, "I'm going to demonstrate a 300-mile-range electric vehicle," and he did. I think the people are more important than the money, but you need some money to do the work. Also, the people who were successful in developing EVs worked on them for a long time. Wally worked on EVs most of his adult life. Alan Cocconi worked many years inventing and perfecting this technology. They understood the problems and worked to solve those problems, which took a long time. I guess dedication can go a long way, but it did all start at Caltech.

ZIERLER: From all of your contributions, from graduate school to professor, what are you most proud of in relation to the development of EV technology?

EDWARDS: When GM turned down the Electro-Spirit proposal, I thought it was kind of the end of EVs. Wally, however, viewed it as the beginning because they had everything they needed, and we kind of helped assemble that. That was kind of the start of the EV development. You had AeroVironment, you had people with power electronics and batteries. Of course, Sunraycer helped bring that together, too. But after Sunraycer, when they got to what they wanted to do next for General Motors, Paul MacCready brought out the Electro-Spirit proposal and said, "Let's go for this."

My understanding is that Alan Cocconi only wanted to work on a high-powered sports car. The ElectroSpirit proposal was for developing a small, four-passenger sedan. That was a big difference from the original proposal, but otherwise it was essentially the same proposal. Sealed lead acid battery, high-power inverter, and an efficient vehicle. And they had the team together. Assembling that proposal and helping to develop the technology for electric vehicles is what I am most proud of. I don't know if it'll be recognized, [Laugh] but we did a lot to move the right technology along to where it could be demonstrated. We worked for seven or eight years trying to put this stuff together so a practical, 100-mile range EV could be made.

ZIERLER: Finally, on that point, looking to the future, best-case scenario, where do you want to see sealed lead acid battery technology adopted, and what's the timeframe that it might happen?

EDWARDS: I think for energy storage. That's what we need right now. Again, you can go with lithium-ion batteries, but they're expensive. With an electric vehicle, you're willing to pay that price because it moves. You've got to carry that around. For utility storage, the weight does not come into it that much. The volume, power, and life are more important, but it should be a lot cheaper, a lot more cost-effective. If you want to do battery energy storage, that would be the place where sealed lead acid batteries might play a role.

There are all kinds of different energy storage systems, and sealed lead acid batteries might be one component in a larger storage system. Our high-power battery can respond very quickly to changes in demand. One of the things we're looking at for fast EV charging is to use high-power sealed lead acid batteries to do battery to battery charging. You charge the sealed lead acid batteries off the grid, then you use those batteries at a station to charge your EV at a very high rate. You'd be continually doing that during the day and recharge at night when the electricity demand is low.

For fleet applications, where you may have these electric vehicles coming at a known time, you could charge the lead acid batteries up at a lower rate and use these batteries to rapidly charge the EVs. The vans or whatever type of fleet vehicles you have could be quickly charged and returned to their route. Those are the types of energy storage systems I see. And there's still a possibility for a limited range electric vehicle just like the proposed Electro-Spirit vehicle where the cost is low.

ZIERLER: It may just happen.

EDWARDS: Yeah, you never know.

ZIERLER: This has been a great conversation. I'm so glad we connected, and I was able to capture your perspective for this. Thank you so much.

[END]

Image Lightbox

Download Full Image

Image Lightbox

Download Full Image

Image Lightbox

Download Full Image

Image Lightbox

Download Full Image

Image Lightbox

Download Full Image

Image Lightbox

Download Full Image

Image Lightbox

Download Full Image

Image Lightbox

Download Full Image

Image Lightbox

Download Full Image

Image Lightbox

Download Full Image