On August 11, 2020, the Transportation Department’s Office of Inspector General announced it would conduct an audit of the National Highway Traffic Safety Administration. The Inspector General cites the fact that 36,000 people died in traffic in the US in 2019 as cause for the audit. It may seem this has no connection to a recent Tesla Model Y accident in which the front passenger seat bent backward. The seat probably complies with FMVSS 207, which establishes requirements for seats, their attachment assemblies, and installation. The problem is this makes no difference to safety.
You may see the Model Y's curved seat and wonder how serious this can be. The answer, simply put, is that it can be fatal. Not only to rear-seat passengers – especially children – but also to people sitting in the front seats, as this video from March 1, 2016, shows.
The video tells the stories of multiple people who were hurt because the seats in their cars were inadequate to protect them. It shows how Taylor Warner and Crystal Butler died because of them. They were 16 months old and 7 years old, respectively, when bent seats put an end to their stories. This places the Model Y case into a different perspective: it is not only a Tesla issue. It is a problem for US consumers as a whole.
George Hetzer III got in touch with InsideEVs after reading our article on the Model Y seat. He is an engineer with a lot of experience in car seat designs, having worked for TS Tech Americas, Honda’s Tier 1 supplier for seats. Hetzer took care of “development, validation, and verification testing/documentation from prototype through mass production and all the milestones in between.”
The engineer says he worked on more than 100 seat projects. If it sounds like too much, especially considering some cars may share seats, that's because “there are export models for each model type." We’ll bring this up at a later point in this article.
Hetzer told us this about the federal safety standard for seats.
“To summarize, FMVSS 207 is a static pull test. A hydraulic cylinder applies a load in the forward, rearward, and longitudinal directions. One thing you’ll notice immediately upon reading FMVSS 207 is how old-fashioned it is. It has test conditions for seats without seatbelts and hasn’t been revised in decades.
Most significant is that FMVSS 207 is a static load test. The result of a static load test tends to have very little correlation with the results of a dynamic strength test (e.g., a car crash that occurs entirely over tens of milliseconds). There are many articles discussing the inadequacies of 207 and proposals for more modern testing requirements. Due to fatal accidents, litigation, and the likelihood a new specification will be implemented in the future, automakers have created their own internal specifications for dynamic seat strength limits.”
We asked Hetzer why he thinks FMVSS 207 still stands as the only standard for seats in the US.
“NHTSA says its ‘5-Star Ratings Program has a limited budget and must concentrate its ratings on front and side-impact crashes that are responsible for the highest percentage of deaths and serious injuries.’ While this is true, it has become a point of contention among automotive safety researchers. A failure mode that has a high probability of mortality/fatality should be addressed regardless of statistical analysis on a large dataset. NHTSA does perform FMVSS 301, which includes a semi-rigid rear impact, to test for fuel spillage.”
FMVSS 301 is another safety standard and you'll want to remember it for later.
A more widespread issue
Hetzer preferred not to comment on the Model Y accident. There is no way to have a professional opinion on that incident without knowing how big the passenger on the cramped seat was, how much the seat was reclined, the speed and weight of the truck that hit it, etc. Jason Levine, executive director of the Center For Auto Safety, said the same, but he did not refrain from criticizing FMVSS 207.
“The fact that Tesla, a company considered by many to be on the cutting edge of fuel and driving technology, could choose to use a seatback that is no more protective in a crash than a lawn chair is the perfect example of the need to upgrade our seat safety standards. No matter how good autonomous driver technology gets over the next few decades, vehicles are likely to crash, and occupant protection will remain a pressing need. It would be nice to see car companies who want to lead us all into the future of not needing a driver to also lead on the need to protect the consumers who they expect to be in their driverless cars.”
If you want to understand the reference to the “lawn chair,” make sure you watch the video below, recorded on May 20, 2010.
This video alone is more than 10 years old, which shows how long the press has been talking about this issue. So has the Center For Auto Safety, which has a dedicated section on its website solely for seat safety and has been trying to change FMVSS 207 for many years now, to no avail.
“We have asked NHTSA to update the standard – often. We were glad to see Senators Markey and Blumenthal introduce legislation last month that would require such an update.”
Levine criticizes the fact this safety standard is more than 50 years old already (it was established in 1966), which would be fine if the requirements were adequate. They’re not, as Hetzer also made clear.
We asked Levine if Europe has a more stringent law regarding seats, for example.
“There is a slightly different standard, but we do not believe it is good enough.”
The Center For Auto Safety executive director refers to ECE 17. The entity explains the discrepancies and similarities between it and FMVSS 207 in the article to which this link leads. The fundamental difference is that Europeans do not rely on these rules for buying cars: they check Euro NCAP tests, according to Alejandro Furas, vice president for Technological Affairs and Secretary-General for Global NCAP.
“In Europe, OICA’s pressure to keep standards as they currently are avoids changes, but Euro NCAP’s pressure on automakers is higher than what the regulations demand. If the rules are not updated, NCAP is a backup in which consumers rely on. All automakers do their best to do well in Euro NCAP.”
In that sense, changes in the ECE regulation are not that urgent, even if desirable. Global NCAP has its own standards for testing the seats.
“Apart from checking the whiplash effect, we make a geometric static evaluation on the seats. If they do not pass it, they are not even submitted to the dynamic tests. In case they make it, we evaluate the seat bending. It cannot go beyond a limited angle. If a seat bends like the one on that Model Y accident, children can hit their heads on the front seat, or the adult’s head can hurt them as well.”
The main video in this article shows what Furas said. In a rear-end crash, we have a phenomenon that is precisely opposite to the submarine effect. Instead of slipping under the seat belt, the passenger is thrown up and rearward if the seatback bends. It is a launching effect, so to speak.
Remember the FMVSS 301 we asked you to keep in mind? Hetzer now explains why it is relevant to this story.
“Going back to FMVSS 301, the dummies are not instrumented, and it is not a test of the seat structure. However, it does give some indication of the likelihood of seatback failure in a rear-end collision. Typically, the NHTSA makes pre/post-test photos and videos available to the public on their online database. Simply searching YouTube for FMVSS 301 will yield all years, makes, and models of vehicles. I spent a considerable amount of time looking through the database and was unable to find a report, photo, or videos for FMVSS Rearward for any Tesla year or model. I would like to know if this test was performed. Perhaps I’m missing something.”
Considering it is meant to measure fuel spillage, we did not expect electric cars to be submitted to it. Yet, we have a video of a Nissan Leaf doing it.
There is one for the Coda and also this one below, which shows a Toyota RAV4 EV.
Did you see how the front seats lean back in these tests? It is so evident that even someone who is not a safety expert would be concerned. Yet, NHTSA seems to have ignored that. The agency could use FMVSS 301 to evaluate how safe these components really are.
Compare the videos below. The first one shows a Ford Fusion test. We chose to start it precisely at the point that presents how much the front seat bends – in slow motion, to make it easier to check.
You can see the dummy’s head almost disappear under the rear door. Now compare that with the Volvo XC90 FMVSS 301 assessment video, also at the slow-motion point.
Why does the European model front seat bend so little while the American one almost goes flat? According to Furas, it has to do with the way these cars are certified.
“The American certification system works with automakers’ declarations of compliance. In other words, it is much more flexible and relies on trust, or else, the government believes in what the manufacturers say. In Europe, validation demands official certifying tests. Any given country leads the process in the European Union, and the tests are paid by the manufacturers. Taxpayers’ money is for essential services. If anyone wants to sell anything to European customers, they are the ones in charge of proving these products are safe, according to the rules. In the US, NHTSA has money to buy a car and test it if it thinks it is necessary.”
We reached out to the agency concerning this “trust” certification system, and this is what NHTSA had to say.
“The Vehicle Safety Acts provides for robust, science-based, and objective performance standards. All manufacturers must certify that their vehicles comply with all applicable Federal motor vehicle safety standards, and ensure that they are free from safety defects before sale. NHTSA purchases a large number of vehicles to test for compliance each year and conducts its own investigations based upon comprehensive mandatory information reporting from manufacturers and tens of thousands of vehicle owner complaints every year. If NHTSA discovers a non-compliance or a defect that creates an unreasonable risk to safety, NHTSA will not hesitate to act, including initiating recalls—which require the manufacturer to refund the customer, replace the vehicle, or remedy the non-compliance or defect without cost to owners of affected vehicles. Note that NHTSA’s defect authority is independent of the existence of any Federal motor vehicle safety standard, and serves as a minimum standard of safety that applies to all vehicles and equipment. In fact, the vast majority of recalls are based upon the existence of safety-related defects. Manufacturers who know or should know of a defect or non-compliance and fail to notify the agency are subject to civil penalties of up to $22,321 per vehicle.”
Furas told us NHTSA has a robust way to deal with safety defects. In a matter of four weeks, it could halt a factory production if it wanted to. That is probably the issue more than anything else: does NHTSA genuinely want to provide safer cars?
If it does, why does it ignore how much the front seats bend in FMVSS 301 tests, even if they weren’t under scrutiny there? Furas mentions something that can explain why the Inspector General decided to audit the agency.
“The US had an important halt in terms of traffic safety about 15 years ago. Europe, Japan, and Australia moved forward. IIHS and NHTSA do not test pedestrian protection, for example. They were going to do that in 2016, but the new administration back then seems to have suspended those plans.”
Cultural differences? Not quite
Global NCAP Secretary-General Furas stresses a relevant difference between the US and European certification systems.
“While NHTSA can enforce a recall if a safety issue emerges, anything similar in Europe immediately stops car sales.”
We have recently seen that happen with the Ford Kuga PHEV. Oddly, the company still did not release more information on what is causing fires in the plug-in hybrids made before June 26, 2020. Furas went on to tell us about how American manufacturers calculate their risk of cost.
“Since automakers cannot afford to stop selling cars, they make a solid effort regarding production consistency in Europe, which includes worker training. That’s cost. In other places, such as Latin America and the US, that does not seem to be a concern. They make a calculation on how much they can save in each car if they don’t do this or that, compare costs, and that’s it.”
Unfortunately, this is common practice in the automotive industry. It is just surprising that it also happens in a developed country such as the US.
Halting car sales in Europe is not something that depends solely on consumer complaints. On top of having a vehicle certified, it is submitted to a Conformity of Production (COP) test every two years after certification to prove it is still compliant. That happens with all cars over there while they're in production.
If you think that is just a matter of cultural differences, think again. Furas pointed out a great comparison between the safety standards that are enforced for cars and airplanes around the world.
“It is ironic that airplanes are certified very much in the same way in Europe and the US, or else, they have to go through official certification tests. When it comes to cars, that is a completely different story.”
Why does the US government trust automakers, but not airplane factories? Is it because car crashes are commonplace – even if they kill so many every year – and airplane crashes are seen as tragedies?
Another misconception that needs to be addressed is that European carmakers make safer cars. What matters is the market where they are made and, sometimes, where they are meant to be sold.
You can see that in these FMVSS 301 tests embedded above and below. The one on the top shows a BMW X3 made in the Spartanburg plant, located in Greer, South Carolina. The one below presents a VW Arteon, a car that is imported from Europe.
Did you see their seats? Generally speaking, if the car is made at or meant for the American market, expect far worse (and cheaper) seats. This is why Hetzer worked on so many projects. Seats for Japanese manufacturers also follow this sad standard, as you can see in the main video in this article and also below, for the 2015 Toyota Camry.
Furas shared an assessment he did on his own.
“I made a test, which has nothing to do with my work at Global NCAP: I analized two BMW 3 Series. Exactly the same specs, same mileage, everything was the same apart from the markets where they were sold. The American one rattled quite a bit. The European car was almost like new.”
Scale is the name of the game… Or isn’t it?
Before you think that Furas is fighting his corner, you’d better know he is from Uruguay. Through Latin NCAP, he has been attempting for many years already to make Latin American cars as safe as the ones sold in developed countries. In that role, he is still fighting to understand what is the rationale behind corporate decisions at automakers.
“If you produce a car that can be sold anywhere in the world, you can order the same parts everywhere. Buying 100,000 components is one thing; purchasing 1,000,000 or more makes each of them much cheaper. Why do manufacturers have a strategy for each region if they sell the same models almost everywhere?”
We also do not have an answer for that. Scale, after all, is the name of the game. Sergio Marchionne dreamed of a merger that would raise FCA’s production to above 6 million cars a year. Carlos Ghosn created Renault’s alliance with Nissan to get there. That said, the fact these companies do not make each and every single factory they run able to supply all markets in which they compete is a mystery.
Bringing that back to Tesla, we’ve already heard the "Made in China" Model 3 is of much better quality than the one produced in the US. When Giga Berlin is working, it will have to follow the rules Furas explained in this article. Why not have a global standard, though, with a better scale of production for components?
Our readers from developing countries may say that’s because adding equipment costs money, but manufacturers could save that money with larger scales of purchasing. What about the workers' training Furas mentioned? Or keeping production consistency? That also costs money. People in these places aren't the only ones driving death traps; the US also needs to improve the safety of cars sold in this country.