NHTSA Adjusts Autonomous Vehicle Rules Ahead Of Tesla Robotaxi Rollout

Sign up for CleanTechnica’s Weekly Substack for Zach and Scott’s in-depth analyses and high level summaries, sign up for our daily newsletter, and/or follow us on Google News!

Last Updated on: 20th May 2025, 02:03 pm

The National Highway Traffic Safety Administration (NHTSA) this week amended its rules relating to autonomous vehicle safety. Previously, the Automated Vehicle Exemption Program allowed imported AVs that did not meet Federal Motor Vehicle Safety Standards to test on US roads. On April 24, 2025, NHTSA issued new guidelines that grant the same exemption from FMVSS compliance to domestically manufactured autonomous vehicle models.

The new policy for autonomous vehicles is intended “to cultivate this Nation’s tremendous potential for automotive innovation,” NHTSA said in a statement about the news rules change. “This change is a significant first step in advancing United States leadership in advanced vehicle technologies; it will not be the last. The agency is actively engaged in developing a multifaceted regulatory framework for ADS-equipped vehicles.”

When announcing the new rules, Transportation Secretary and former Faux News on-air personality Sean Duffy said, “this administration understands that we’re in a race with China to out-innovate. Our new framework will slash red tape and move us closer to a single national standard that spurs innovation and prioritizes safety.” The new standards will enable the “safe commercial deployments of AVs.”

Car and Driver notes the new rules package will focus on robotaxis for ride hailing services, and do not apply to automated driving systems used in privately owned passenger vehicles. “The new exemption program rules will make it easier for Tesla to deploy the U.S.-made Cybercab, since {its} lack of physical controls means it doesn’t meet the FMVSS rules,” it says.

Protecting Autonomous Vehicle Manufacturers

Wired reports that although the new policy shields manufacturers from disclosing some information about how their automated driving systems operate in the the real world, it does eliminate some redundancies that made the data more difficult to understand and handle. Nevertheless, companies that deploy self-driving cars must still report crash information to the federal government.

Noah Goodall, an independent researcher who studies autonomous vehicles, told Wired the changes may make it harder for outsiders to spot or understand patterns in the errors self-driving vehicles make. He did acknowledge, however, that the public database on crashes has been difficult to work with since it was launched in 2021. “You’re getting less reporting now. From my perspective, more data is good,” he said.

The changes will limit “one of the only sources of publicly available data that we have on incidents involving Level 2 systems,” says Sam Abuelsamid, a writer who covers autonomous vehicle systems. For instance, the public will no longer be able to see what version of the software was in use when reportable events occur, something companies consider “confidential business information.”

Minor incidents like fender benders that result in less than $1000 in damage will no longer need to be reported. Makers of automated driver assistance systems (ADAS) such as Tesla’s Full Self-Driving are now exempt from reporting crashes that do not result in fatalities, hospitalizations, air bag deployments, or involve a “vulnerable road user,” such as a pedestrian or bicyclist.

“This does seem to close the door on a huge number of additional reports,” William Wallace, the head of safety advocacy for Consumer Reports told Wired. “It’s a big carve-out.” The changes are contrary to CR’s advocacy for federal rules that address  “significant incident underreporting” requirement for makers of advanced vehicle technologies. “The company that probably benefits the most from that is Tesla,” Abuelsamid said. Though the Transportation Department cited safety as the number one motivator behind the new rules, “there’s nothing in these changes that actually prioritizes safety,” he added.

Preparing For The Cybercab

Tesla recently unveiled a prototype of what it calls its Cybercab, a two-passenger vehicle with scissor doors similar to those that debuted on the Lamborghini Countach in 1974, considered ultra cool by car customizers. Perhaps the most notable feature of the Cybercab is the absence of a steering wheel, accelerator, or brake pedal. It clearly does not meet FMVSS rules, but the new policy will exempt if from that requirement.

Elon Musk has been telling anyone who will listen that he and his team are “laser focused” on starting a ride-hailing service in Austin, Texas in June. That experiment will not use Cybercabs but rather Tesla Model Y SUVs that will be modified to use Full Self Driving Unsupervised software. Current Tesla owners have advanced versions of FSD Supervised.

On May 16, 2025, Morgan Stanley analyst Adam Jonas wrote that the launch of the robotaxi service in Austin was “a go,” but with significant limitations. “Think 10 to 20 cars. Public roads. Invite only. Plenty of tele-ops to ensure safety levels (‘we can’t screw up’). Still waiting on a date,” he said. Investors Business Daily reports that Musk said recently that “unsupervised autonomy will first be solved for the Model Y in Austin.” Some have observed if there is a team of humans monitoring every action those cars take remotely, calling it “unsupervised autonomy” may be a bit of a stretch.

The Robotaxi Test

Recently, Lloyd Lee and Alistair Barr of Business Insider conducted their own test of robotaxi technology by taking a Waymo autonomous taxi ride and a ride in a Tesla with FSD from San Francisco’s famous Twin Peaks to Chase Center near Mission Bay — a distance of 4 to 7 miles, depending on the route chosen.

“We chose these destinations for two reasons. One, it would take the cars through winding roads and both suburban and city landscapes. And two, there were a few ways to get to Chase Center from Twin Peaks, including the 280 highway. Waymo’s robotaxis can’t take riders on the highway yet. Tesla can,” the pair said in a report for Business Insider that was republished by AOL.

First the pair made the trip in a Waymo Jaguar I-Pace equipped with Version 5 of the company’s Drover software suite, which includes five lidar sensors, six radars, and 29 cameras. “Throughout our ride in the Waymo, we were impressed by the AI driver’s ability to be safe but assertive,” the pair wrote. “The Waymo was not shy about making yellow lights, for example, but it never made maneuvers you wouldn’t want a robot driver you’re entrusting your life with to make.”

They report the Waymo car behaved like a human driver trying to make the trip in the shortest possible time. At one point, the car came up behind another car at a stop sign and switched to the vacant lane to the right in order to move ahead of the slowpoke it had been behind. Most of us would do much the same, given the opportunity.

Next, the duo made the same trip in Barr’s 2024 Tesla Model 3 equipped with Hardware 4 and FSD Supervised software v13.2.8. They note that Tesla released a minor update to the software a few days after their test was completed. Hardware 4 includes eight external cameras. “It should be noted that this is not the same software Tesla plans to use in the robotaxis set to roll out this summer. The company said it plans to release FSD Unsupervised, a self-driving system that will not require a human behind the wheel,” they noted.

So, how did the Tesla do? Pretty darn well, especially considering all the extra sensing power available to the Waymo car. All in all, the pair rated the performance of the Tesla as nearly equal to that of the Waymo — until it ran a red light. The intrepid testers describe what happened this way:

“The incident occurred at a fairly complex intersection that resembles a slip-lane intersection, but with a traffic light. The Tesla’s console screen showed how the car detected the red light and came to a dutiful stop. Then, despite the traffic light not changing, the Tesla drove ahead.

“We didn’t come close to hitting any cars or humans on the street…. However, the vehicle slowly drove through this red light, which left us both somewhat shocked at the time. Some Tesla drivers appeared to have reported similar issues in online forums and in videos that showed the vehicle recognizing the red light but driving ahead. One YouTuber showed how the Tesla first came to a stop at a red light and then continued driving before the light changed. It’s unclear how common this issue is. Tesla hasn’t publicly addressed the problem.”

One other glitch happened along the way. San Francisco has many bike lanes that are painted bright green. At one point in the journey, the Tesla wandered into one of those lanes and drove along for several seconds before returning to its proper lane. Neither incident will need to be reported to federal authorities according to the new rules that Sean Duffy claims prioritize safety. CleanTechnica readers are free to make up their own minds about how and why the change was made by Duffy and his minions.

The concern many readers have is that a Tesla in robotaxi service looks the same as any other Tesla. The Waymo is festooned with pods, modules, and bulges that make it look like it wandered away from the set of Lost In Space. It’s hard for drivers not to notice a Waymo autonomous car in the vicinity. A Tesla robotaxi has very few external clues to alert drivers to its special status.

The government’s new fixation on “safety” seems more of an attempt to protect a staunch supporter of the current president than an honest attempt to protect the safety of the public. Less information is seldom in the public interest.