American roads looked different when the Insurance Institute for Highway Safety built its car-smashing crash test system in the early 1990s. Backup cameras didn’t exist, airbags weren’t mandatory, and safety rules had not yet killed the pop-up headlight. But perhaps the biggest difference is that cars were a lot less heavy—about a quarter lighter than the average vehicle today.
Raul Arbelaez, who oversees crash tests at the IIHS, has watched that transformation occur over his two-decade career, seeing sedans morph into “crossovers” and minivans become SUVs. The nonprofit lab tests most of the popular cars on the market and its safety endorsements are coveted by automakers. But until recently, he didn’t have a reason to question whether the IIHS equipment would be up to the task of crashing heavier vehicles. That reason was the Hummer EV.
At more than 9,000 pounds (4,000 kilograms), the electric SUV is about a third heavier than the weightiest vehicle IIHS has ever tested—which happens to be another EV, the Rivian R1T—and more than twice as heavy as the average American car, which weighs about 4,000 pounds. So Arbelaez bought a few cheap, old pickup trucks and began loading them up with concrete to match the Hummer’s weight. Despite the extra strain on the cables, the system held. The pickup went boom and a video of the test had a viral moment.
But Arbelaez is still worried. The trucks and SUVs that have taken over US roads in the past two decades protect their own passengers exceptionally well. But many of the same qualities that help a vehicle’s safety rating—including increased frame stiffness, size, and weight—also make them more of a menace to everyone else. Vehicle safety standards mostly reflect the safety of people inside a vehicle, not those outside. And despite improvements in safety technology and car design, heavier vehicles have contributed to increasing road deaths in recent years, according to data from the US National Highway Traffic Safety Administration—particularly among pedestrians, cyclists, and people driving smaller vehicles. Now EVs are tipping the scales even further. They often combine the beefy dimensions of supersize SUVs with a battery that itself weighs as much as a small sedan. Oh, and also blazing acceleration.
When WIRED put that to General Motors, spokesperson Mikhael Farah pointed to data linking pedestrian deaths to factors such as worsening pedestrian infrastructure and more speeding and drunk driving. (The same research noted that deaths due to SUVs were increasing faster than those due to cars, and that crashes involving heavier vehicles are generally more severe.) The company also pointed to safety features of the Hummer, such as upsized brakes and crash avoidance systems.
But Arbelaez is not alone in his concerns. Last month, in a speech to fellow safety experts, Jennifer Homendy, head of the US National Transportation Safety Board, made special mention of the “unintended consequences” of EVs—not just Hummers, but also electric Volvos, Fords, and Toyotas that carry thousands of pounds of extra weight beyond conventional cars of a similar size. “That has a significant impact on safety for all road users,” she said.
It’s a tricky balance. Both Homendy and Arbelaez say they are excited about electric cars and care deeply about addressing climate change. But unless automakers or regulators find ways to reduce vehicle weights, they also fear a knock-on consequence of size inflation will continue unabated: that the only way to feel safe on roads with massive cars is to barricade yourself in one of your own.
WIRED: So you’ve been crashing some really heavy vehicles. Why?
Raul Arbelaez: We’ve been conducting tests on electric vehicles for several years now, but I wanted to be sure that we’re prepared to test something this heavy. So I said, “Can you get me an old junker?” Some older model pickup. Some of the Youtube commenters weren’t too happy that I was calling them junkers. They said, “Hey, I’d fix them up.” But we got them for close to nothing from the junkyard. All I needed to do was hook it up to our propulsion system and figure out if it can get up to our highest crash speed, 40 mph for our frontal crash. My biggest concern was that we wouldn’t have enough energy to pull the system. Our system has a series of 16 cylinders that have nitrogen in them, which is compressed to store energy. It typically takes up to four minutes to charge and store that energy. For this test, it approached 10 minutes.
So it worked. And at some point, whether it’s the Hummer or some other EV truck, you’re going to be testing vehicles this heavy. What do you expect from that?
What happened over many years with improvements in powertrain technology is they didn’t always go toward making more efficient engines. They went into making more powerful, higher-torque engines that are great for marketing but maybe aren’t the safest thing for people.
With electrification, I hope we can change course. I hope we can say, you know what, we’re not going to make this thing 9,500 pounds. We can shave a few thousand pounds off and not get 1,000 horsepower and 300 to 400 miles range and that’s fine. Maybe instead let’s focus on infrastructure improvements that make it easier for people to charge more often and get rid of this range anxiety. And can we put in a battery that’s half the mass and save some of those resources and maybe make the vehicle more nimble. But it certainly seems like instead the momentum is in favor of heavier and stiffer cars.
Stiffer too? What does that mean in a crash?
A designer of a heavier vehicle has two choices for managing the extra energy of a crash. The first choice is to make the front-end much longer. There is a little flexibility there, but there are limits. The other option is to increase the stiffness of the front-end. That makes it more dangerous for every other smaller vehicle. You’ll have more people who are vulnerable. All of a sudden that 20 mph crash where the other car was safe isn’t safe anymore.
Just like anything else in safety, there are tradeoffs. You have to design a vehicle in two ways: enough stiffness to handle your own vehicle’s mass in a single-vehicle crash, but not so much it’s a battering ram to all other vehicles. Something like a larger pickup has to be stiffer to protect the occupants in the vehicle. But once it hits anything smaller it becomes potentially more dangerous. A lot of the time the path to doing the right thing isn’t obvious.
What does that look like in practice?.
One thing that has been known for a long time is that when cars crash into each other you want them to ideally be nominally the same weight, so you don’t have the equivalent of a 350-pound NFL linebacker running into a 100-pound high school freshman. If they’re going the same speed, the high schooler will always lose, and so will the smaller car. It’s the law of conservation of momentum. If I crash two 4,000-pound Honda Pilots into each other at 40 mph, they will go to 0 mph. Say that instead one of those vehicles is a 9,500 pound Hummer. The Pilot, a pretty big SUV, starts traveling back at 18 mph. So it’s basically gone from a 40 mph crash to 58 mph. That is a whole lot more energy that goes into that vehicle. The rating for the Pilot in that crash would go from good to poor.
That’s just for a Hummer hitting a Pilot, which is a pretty big SUV. The other type of compatibility that we want to make sure of is that the structures align for a front-to-front crash or even a front-to-side crash. You don’t want an arena monster truck that is 5 feet off the ground running over all the other vehicles. You want the bumper bars and the frame rails to be in the same nominal zone so that safety structures can absorb parts of the crash.
It sounds like the crash test system incentivizes people to purchase bigger cars in order to be safe. The bigger you are, the safer you and your passengers are. But then you become more of a menace to the people and cars around you.
I was the lead engineer back in 2003 when we developed the original side crash test. Some of the first research tests were striking cars with 1,900 kilograms, but we found that vehicles at the time were just so bad. They lacked the vehicle side structure and few had side airbags. We compromised and went to 1,500 kg, much lighter than the real weights of SUVs, because everybody would have been poor across the board from the beginning. Even that was considered crazy.
It was a lot more severe than what the federal government was doing. I remember sharing the early development of that test, and this one researcher from NHTSA [National Highway Traffic Safety Administration] got up and asked if we were trying to cause a “mass war.” I thought, oh no, are we really creating this mass war where we make the test reflect the real world with heavier cars and everyone just has to get heavier to perform well? I really hope not. We recently increased the test back to 1,900 kg.
We aren’t out there to force everyone into big vehicles or to small vehicles. Whenever gas prices go up and people buy microcars, we just present the real-world data showing the increased risk of driving those vehicles. We don’t put a Ford F-150 next to a Honda Civic. We put them in the categories we think people are shopping for, based on footprint and mass. And we tell people the safest vehicle within the class they’re considering. Is that system perfect? We’re keeping our eyes open and looking at how things could change.
On our way to going green there are going to be some unintended consequences that we just have to face up to. Somebody is going to have to decide what is acceptable or not. And because electric vehicles are so expensive, we’re going to have some equity issues between people who can afford EVs and people who are in old combustion vehicles.
When are we going to see this in the data? So far, most of the discussion is about how safe EVs are for passengers.
I think it’s going to take a lot more time to see the impacts of heavy EVs. Not every severe crash is investigated in the real world. Even for fatal crashes you don’t know much about the actual crash.
Something that we do constantly is circle back to the real-world data, and the past few times we’ve done it we’ve identified new crash modes to add. We’ve added rear impacts. We’ve added a rear passenger, a small woman to an adolescent child. Unfortunately, in order to do that analysis, a lot of people have to be in crashes. It just takes time.
Anything to be optimistic about?
One thing I am hopeful for is that vehicle crashworthiness or self-protection isn’t the only tool in our toolbox. With the emergence of crash-avoidance technologies—cameras and radar systems that help detect and prevent or reduce the severity of collisions—I think there’s huge potential to drop the total number of severe crashes and crashes with fatalities. I am hoping that the advances made through technology outpace human stupidity.
Nothing drives me more crazy than continuing to see the number of people killed unbelted. We have over 90 percent seatbelt use in the US and yet those unbelted represent 50 percent of occupant fatalities. I hope we can reverse this awful trend from the past few years of more severe accidents on the roads, and a horrible increase in pedestrian fatalities. It’s why I’ve continued to do this job. I think we can make it safer.