Future Tense

When Computers on Wheels Become Flaming Metal-Oxide Robocars

A close-up of the Tesla logo on the front of a red car.
A Tesla Model S P85D in Palo Alto, California on April 30, 2015. JOSH EDELSON/Getty Images

Over the weekend, we learned once again that when a Tesla crashes, it is like no other vehicle in an accident. The facts are still trickling in, but what is known is that two middle-aged men died. And according to reports, they somehow set their Model S into motion without a driver, crashing it into a tree at high speed and causing it to explode in flames that took up to four hours to extinguish. Much of the discussion around this strange incident focused on Tesla’s Autopilot, a feature Tesla claims is intended only to assist the driver but that critics claim is an oxymoron in the automobile context. After all, the word autopilot connotes sustained fully automatic operation with the human operator out of the loop, an operating mode hitherto used only in aircraft and spacecraft. Tesla Autopilot has been implicated in a number of other fatal crashes of Tesla cars, sparking debate between those who blame technology and those who blame the operator. But that’s a misleading distinction, because accidents invariably stem from an intersection of human and non-human factors. This latest incident shows that when things go awry, ideas and assumptions about the capabilities of technology are often as important as the actual properties of the hardware itself.

Advertisement
Advertisement
Advertisement
Advertisement

Tesla made its name by building electric cars around the most powerful battery packs ever devised, composed of cells made of lithium metal oxides. These compounds make for the most energetic rechargeable batteries—but they are also highly volatile. If damaged, cells in lithium batteries can ignite in a chain reaction called thermal runaway, generating fires fed by the oxygen in the metal oxides that are difficult to quench. Since their introduction in the early 1990s, lithium cells in consumer electronics have been involved in string of conflagrations, most notably in Samsung’s Galaxy Note 7 in 2016, a fiasco that destroyed the rollout of this smartphone at the cost of billions of dollars.

The potential for disaster is even greater in large lithium battery packs for electric cars, which contain far more reactive material than consumer electronics cells. Fortunately, incidents are relatively rare, and lithium power is widely recognized as key to the revolution in handheld computing from the 1990s and the resurgence of the battery electric car from the late aughts. Giant battery packs allow Tesla cars to achieve unprecedented performance, not only in terms of range, long seen as the Achilles’ heel of electric cars, but in terms of torque and acceleration. Contemporary electrics are technological marvels, and Tesla is rightly recognized for doing more than any single automaker to usher in the age of zero-emission automobility.

Advertisement
Advertisement
Advertisement

But Tesla has also cultivated a mindset about cars, drivers, and driving that can obscure the complexities of its vehicles. Under the leadership of Elon Musk, Tesla built its brand identity around the idea of an infallible electric supercar, analogizing it as a computer on wheels. Tesla deployed this trope partly to distinguish itself from its rust-belt competitors, who won infamy by canceling the original generation of contemporary electric cars in the early 2000s. But the car-as-computer-on-wheels was more than clever marketing. Tesla tried to apply this idea as a literal engineering precept, equipping its vehicles with massive computing power and downloadable updates for applications including Autopilot.

Advertisement
Advertisement

The resulting cars embody a mix of appeals to difference audiences. For Tesla, the quality of zero emissions is an important but by no means the only desirable feature of electric propulsion. The company has emphasized power and style, framing its vehicles as viscerally exciting personal experiences for drivers. In its entertainment and driver-assist features, on the other hand, Tesla acknowledges that driving can also be tedious and suggests there are circumstances where control may safely be transferred from the operator to the car-computer.

Advertisement

This blend of sophisticated technology can enable a variety of behaviors. In a Tesla, the individual is empowered both by driving the car and sometimes by not driving the car and allowing the computer to do it. In not-so-subtle ways, Tesla also encourages users to take calculated risks. Some versions of the Model S, the sedan that made Tesla a household name, are equipped with “insane” and “ludicrous” modes that turn the cars into electric dragsters. A Tesla stunting culture has emerged, reinforced by association with Musk’s activities in spaceflight, an enterprise whose culture has historically accepted and even valorized risk. Musk’s elision of cars and space pyrotechnics and descriptions of Autopilot as an experimental but necessary safety feature of the computer on wheels have likely given suggestible consumers a misguided impression of what Tesla’s powerful vehicles are capable of.

Advertisement
Advertisement
Advertisement
Advertisement

The broader context is that risk has always been an intrinsic part of automobile culture and Tesla is far from alone in the industry in promoting kinetic thrills. Prior to this most recent crash, Musk responded to criticism of lithium power by pointing to the dangers of gasoline. It is a reasonable observation, but framing safety as a contest between propulsion systems misses a broader point. Most of the thousands of people killed annually in automobile accidents perished as a result of operator error, especially distracted and drunk driving, and the shift from gasoline to electric propulsion is not likely to change that, regardless of what other benefits may accrue from the electric car. To some critics, it is the automobile system itself that is the real problem.

Advertisement
Advertisement

By themselves, lithium batteries and driving aids do not cause accidents. Technologies frequently fail, to be sure, often with fatal consequences, yet failures trace ultimately to human actions. Manufacturers and consumers are engaged in a vast tacit experiment that feeds back knowledge of how products behave in the real world, with recalls being a key way of building knowledge. Over the years, this process has helped make automobile technology increasingly reliable, leading to a corresponding decline in deaths and injuries. Most of the time, where cars are concerned, the judgment of the operator is the difference between routine motoring and a Viking funeral.

Future Tense is a partnership of Slate, New America, and Arizona State University that examines emerging technologies, public policy, and society.

Advertisement