Materially Different

In “Computers on Wheels?” (Issues, Winter 2023), Matthew Eisler makes a significant contribution in understanding the roots of the modern electric vehicle (EV) revolution. He provides many missing details of the thinking behind Tesla’s beginnings, especially the ideas for framing automobiles as consumer commodities. More importantly, he highlights the incompleteness of the “computer on wheels” analogy to the bane of legacy automakers and policymakers alike.

As Eisler notes, while electric vehicles and computers are similar in some respects, they “are significantly different in terms of scale, complexity, and, importantly, lifecycle.” One such difference is the intense demand EVs place on being able to develop and sustain extremely complex software not only for safety-critical battery management but for the rest of the vehicle’s systems. Tesla’s organic software-development capability is a critical reason it has been able to forge ahead of legacy automakers in terms of both features and manufacturing costs. While EV batteries contribute to some 35–40% of an EV’s manufacturing cost, vehicle development costs attributable to software are rapidly approaching 50%.

Although the amount of software reinforces the analogy of an EV being a computer on wheels, the analogy fails to account for how EVs materially differ from their internal combustion engine counterparts. EVs represent a new class of cyber-physical system, one that dynamically interacts with and affects its environment in novel ways. For instance, EVs with their software-controlled electric motors no longer need physical linkages to steer or apply power—a joystick or another computer will do. With additional devices to sense the outside world along with requisite computing capability, EVs can more easily drive themselves sans human interaction than can combustion-powered vehicles. Tesla realized this early and made creating autonomous driving capability a priority. In developing self-driving, the company further increased its software prowess over legacy automakers.

As Eisler notes, policymakers wholeheartedly embraced EVs, first to fight pollution and later to combat climate change. However, policymakers have also embraced the potential of autonomous-driving EVs and are counting on them to limit individual vehicle ownership, thus reducing traffic congestion and ultimately reducing greenhouse gas emissions by up to 80% by 2050. Even for Tesla, creating fully self-driving vehicles has been much more difficult than it imagined, illustrating the dangers of policymakers adopting nascent technologies as a future given.

This highlights another critical problem that Eisler pinpoints as resulting from policymakers’ embracing EVs as computers on wheels—that of scale. Transitioning to EVs at scale not only demands radical transformations in automakers’ global logistical supply chains but also establishes new interdependencies on systems and their capabilities outside their control, from lithium mines to the electrical grid. The grid, for example, will need increased energy generation capacity as well as a significantly improved software capability to keep local utilities from experiencing blackouts as millions of EVs charge concurrently. Policymakers are only now coming to terms with the plethora of network effects EVs, and their related policies, create.

Eisler clearly underscores the myriad challenges EVs present. How well they will be met is an open question.

Robert N. Charette

In 2019, I faced the chore of replacing the family car. In visiting various dealerships, I found myself listening to lengthy descriptions of vehicle control panels, self-parking features, integration of contact lists with built-in phone systems, and navigation and mapping options. I heard nothing about safety features, engine integrity and maintenance, handling on the road, passenger comfort, or even gas mileage (I was looking at all types of engine options). I was barely even encouraged to take a test drive. It was as if, indeed, I was shopping for a computing machine on wheels.

By opening with the “computer on wheels” metaphor, Matthew Eisler provides an opportunity to think about a major technology shift—from the internal combustion engine to the electric motor—from multiple perspectives. How is a car like a computer? How does a computer operate a car? How did electric-vehicle visionaries adapt design and production techniques from the unlike business of producing computers? How are the specific requirements of the single-owner mode of transportation different from other engineering challenges? How are geopolitical crises and the loci of manufacturing of component parts implicated in car production? What might answers to these questions tell us about complex systems and technological change over time?

As Eisler deftly argues, the modern push for electric cars represented the confluence of multiple social, economic, technological, and political scenarios. EV enthusiasts looked outside Detroit for new approaches to car building. The methodology of the information technology industries offered the notion of assembling off-the-shelf component parts, but the specific safety requirements—and the related engineering complexity—of automobiles put the process on a longer trajectory. On the one hand, the near simultaneous cancellation of General Motors’ early electric car EV1 and bursting of the dot-com bubble discouraged investment in a new EV industry. On the other, public sentiment and resulting public policy created a regulatory environment in which a technically successful EV could flourish.

Eisler highlights additional tensions. A fundamental mismatch between battery life and engine life undermined the interest of the traditional auto industry in these newly designed vehicles and belied difficulties in production and maintenance. And the trend of globalization, with its attendant divide between design in the West and manufacturing in the East, persisted beyond policy initiatives to onshore all elements of car production in the United States. Most profoundly, taking the longer view toward the future, Eisler indicates that thinking about cars as computers on wheels fails to consider the larger sociotechnical system in which EVs are inevitably embedded: electric power networks.

Today, Americans plug mobile computing devices without wheels into outlets everywhere, expecting only to withdraw energy. Sure, some of us recharge our phones with laptops as well. But we don’t really see laptops as storage batteries for the power grid. Nor do we generally consider when to recharge phones to avoid periods of peak power demand. But the energy exchange potential of an all-EV fleet that may replace current hydrocarbon-burning cars, buses, trucks, and trains suggests a much more complex electrical future. Eisler gives just one of multiple examples, noting that EV recharging shortens the service life of local power transformers. EVs are complicating construction and maintenance of power distribution networks and are already much more than computers on wheels. A wide range of industries have difficult and interesting questions ahead about whether and how these hybrid IT/mobility devices will fit into our highly electrified future.

Julie A. Cohn

Matthew Eisler makes a welcome contribution to the emerging conversation about the history of the rebirth of the electric vehicle. In particular, Eisler rightly highlights two factors that are often left out of breathier, more presentist accounts.

First, Tesla—the company synonymous with the faster, better EV that competes head-to-head with internal combustion—was not founded by Elon Musk. Although Musk subsequently negotiated a deal by which he was officially recognized as a Tesla founder, Eisler rightly focuses on the efforts of Martin Eberhard and Marc Tarpenning (and later JB Straubel) who recognized the opportunity arising from the confluence of advances in lithium-ion (Li-ion) storage batteries and drivetrain technology. Although many liken Musk to an “electric Henry Ford,” it is a poor analogy, as Eisler makes clear.

Second, Eisler rightly focuses on the social contexts of innovation in both the consumer electronics and automotive industries. The story of the Tesla founders trying to convince a reluctant Asian manufacturer to sell them high-performance Li-ion batteries for their pilot vehicle (eventually the Roadster) stands in sharp contrast with current efforts to “blitzscale” battery production for the growing electric vehicle market. The early battery makers focused on consumer electronics and therefore underestimated demand for Li-ion cells from electric vehicle manufacturers. Conversely, today’s rush to mass produce Li-ion cells everywhere may lead to overinvestment and rapid commodification rather than future riches. The Li-ion-powered electric vehicle is an industrial accident, not a carefully orchestrated transitional technology. Its history is definitely not one characterized by the seamless adjustments of efficient markets, a point further underscored by Eisler’s recognition of the role of state and federal policymakers, both in the initial rebirth of the EV and in support of Tesla.

There are three areas where I think Eisler might have missed the mark:

First, the idea of the car as a computer on wheels goes back to the dawn of the solid state era. In my own work on the history of EVs (The Electric Vehicle and the Burden of History, Rutgers University Press, 2000), I found engineers in Santa Clara county in the mid-1960s talking about the electrification of the automobile as a result of advances in solid state electronics. But it turned out that the incumbent auto industry responded by electrifying everything except the drivetrain. The car was an “electrical cabinet on wheels” before it became a computer. In this respect, thinking about electrification predates the birth of Silicon Valley, not to mention the dot-com era and everything that followed.

Second, many historians of technology may not wish to hear it in such stark terms, but it is very hard to imagine the EV transformation Eisler describes occurring in the absence of the important technological advances in energy storage and drivetrains. The “computer on wheels” was simply not plausible in the late-1980s. Innovation matters. Technological change creates affordances that shape downstream social and economic outcomes. Before those affordances were available, much of Eisler’s story would not have been possible. Third, the success of the standalone electric vehicle may have blinded Eisler (and others) to some of the paths not taken. For many years, EV supporters focused less on the electric vehicle as a replacement for internal combustion and more on adjacent market opportunities. For this group, electrification might have looked like electric scooters or electric-assist bicycles, or like micro cars such as city- or neighborhood-electric vehicles, or even like electric light delivery vans and small buses. Recent events have pared away the many other ways that the electrification of the auto might have developed.

David A. Kirsch