AS PREPARED FOR DELIVERY
Thank you, Director Kratsios. And again, thanks to Secretary Duffy.
It’s a good measure of this Administration’s focus on AVs that we got two such esteemed Administration leaders with us today. And I know that support goes all the way to the top.
I want to build on Secretary Duffy’s remarks by taking a double-click look at where we stand and where we’re headed on AVs. As Secretary Duffy said, we cannot afford to move too fast, but we certainly cannot afford to move too slowly. We are moving with a sense of deliberative urgency—enabling technological advancement while maintaining rigorous oversight.
A little background on the reality of NHTSA’s authorities in this space. There is a common misconception that the NHTSA Administrator has a magic wand that can instantly fix every policy or technical challenge. I can tell you firsthand: that isn’t the case. I asked.
NHTSA operates under the Vehicle Safety Act which gives us two primary jobs:
- Create and enforce minimum vehicle or equipment performance standards, known as Federal Motor Vehicle Safety Standards (FMVSS), and
- Provide broad industry oversight around vehicle or equipment defects in design, construction, or performance that pose an unreasonable risk to safety—even if a specific FMVSS does not apply to the technology.
Over the past fifty years, NHTSA has established dozens of FMVSS covering everything from brakes and tires to headlights and windshields. They were meant to be technology-neutral, but were created in the context of the technology that existed when drafted.
Over those years, what was the one thing all vehicles had in common? A human driver. Accordingly, our FMVSS are built around the needs of human operators and require equipment to ensure vehicles meet performance requirements based around human inputs. For example, our light duty braking performance requirements focus on stopping distance after specific forces are applied to a requisite physical foot pedal.
This creates a paradox: The technology of the future is often tethered to the regulations of the past. And this isn’t the case because doing so provides safety benefits—a vehicle designed never to be operated by a human doesn’t get safer because it has a foot pedal—but because the requirements are the requirements, and changing them is difficult. And while NHTSA does have limited authority to grant exemptions for new technologies, reliance upon the generosity of a regulator is a poor business model.
But this is an administration that is looking to enable innovation while performing robust safety oversight — which is the way the Vehicle Safety Act was designed to operate.
We’re looking to support innovation because automated driving systems hold incredible potential promise, unlocking mobility options for senior citizens, people with disabilities, and others who are unable to or don’t want to drive for a variety of reasons—because they’ve been drinking or maybe just because they’ve had a bad day and a long flight and they just don’t want to talk to a stranger. Importantly, this technology has the potential to enable full participation in our society and economy for the many left behind due to an inability to drive.
Because access to transportation is access to freedom.
Additionally, AVs could play a role in addressing the fatality crisis on our roads. Several risky and unacceptably common driving behaviors like speeding or driving distracted, fatigued, drunk or high play a role in tens of thousands of deaths every year.
AVs could potentially eliminate so many of these dangerous behaviors. AVs will never drive impaired, tired or distracted—and they shouldn’t speed through school or work zones. If they do, they should be subject to traffic enforcement. And you don’t need to worry about them having road rage.
But let’s be honest with ourselves: Building a safe and robust automated driving system is one of, if not the, most challenging engineering problems humanity has ever attempted. And it’s safety critical. Think of the infinite possible challenges presented by what happens on our roads, and how the human mind is able to adapt to solve those challenges in real time, when they are attentive and sober.
Automated driving systems have to be capable of handling each such challenge, requiring massive amount of data for training, high fidelity sensor inputs for operations, and an enormous amount of computing power to process it all, in less than a blink of an eye.
Because this is so difficult, we occasionally see odd and annoying behaviors such as vehicles stopping in-lane for no apparent reason. Sometimes, we see genuinely unsafe behaviors requiring NHTSA to use our enforcement authority. NHTSA has not been—and will not be—shy about using our enforcement power. We have launched a number of defect investigations into AV developers and have pushed for and overseen several recalls. And we will continue to do this because public trust depends on public safety. And this requires a cop on the beat.
But the potential promise this technology can provide to society in terms of mobility and safety is far too great to ignore, or worse, discourage or prohibit.
This is why, during President Trump’s first term, we were quite active in the AV space, laying the groundwork with early guidance and regulatory updates. This is also why, after a few years of inactivity, we are back at work.
Under the direction of President Trump and Secretary Duffy, we continue to build out and implement a new Federal Automated Vehicle Framework, based upon three pillars:
- First, prioritize the safety of ongoing AV operations on public roads.
- Second, unleash innovation by removing unnecessary regulatory barriers to the safe development of the technology.
- Third, enable the commercial deployment of AVs to enhance safety and mobility for the American public.
We’ve made strides in delivering on these priorities in the last year by prioritizing reporting requirements to focus on safety, expanding access to our exemption programs, and hosting a collaborative discussion with industry and safety experts on many pressing AV policy issues. And we continue to move forward on these pillars.
FMVSS:
I've discussed previously that we’re evaluating all of the FMVSS for unintended barriers to new designs enabled by automated driving systems, and seeking to amend them in a way that does not have adverse safety impacts. Doing so doesn’t merely enable innovation, it’s just plain good government. Our first tranche of amendments will address FMVSS No. 102, relating to transmission shifting; FMVSS No. 103, windshield defrosting and defogging systems; and FMVSS No. 104, windshield wiping and washing systems.
As a “car nerd,” I could talk about windshield wiping systems all day, but my team tells me that might be going a bit too far for this morning. We’ll save that for the next event—but I’m moderating a fireside chat with AV company CEOs immediately after this and want them to know that wipers are fair game for questions.
We have many more proposals in the pipeline that we’re working with a tremendous sense of urgency to move forward. Keep watching this space.
Zoox:
As the Secretary mentioned, today we’re announcing that NHTSA is officially seeking public comment on Zoox’s request for exemptions from several FMVSS to enable commercial deployment of a purpose-built robotaxi. The Zoox vehicle is designed without manual controls or mirrors, provides for a “campfire” seating arrangement instead of a traditional forward-facing cabin configuration, and is designed to operate bidirectionally.
If granted, this would be the nation’s first-ever deployment exemption provided for a novel-designed AV for passenger service. To grant this exemption, NHTSA must find that doing so is in the public’s interest. It must also find that the vehicle is at least as safe as a vehicle meeting FMVSS.
And we don’t take this request lightly, which is why we want to hear from the public to inform us how we should evaluate this. Our request for comment document was made public this morning, and will be published in the Federal Register tomorrow morning. This includes a link to the docket, where you can review Zoox’s petition and supplemental information and arguments why we should grant their exemption. We’ve heard from them, and now we want to hear from you. Uniquely, attendees will have the opportunity to check it out.
Guidance:
We’re also moving ahead today with developing new guidance for AV developers on AV safety. This will be the first major guidance from NHTSA to AV developers since the AV 2.0 guidance document in 2017. A great deal has changed since then and we know a lot more about technology and safety related to AVs. Just as AV 2.0 helped set a baseline for companies, communities, law enforcement, and safety experts to work from for the last eight years, we expect the next set of guidance to shape the industry’s approach to safety for years to come.
But guidance isn’t something that’s best handed down from on high. We’ve certainly got ideas and strong opinions on safety assurance, but this is best informed by participation from other safety and technical experts. This guidance will be coming out in batches over the coming months, and we’re starting the public aspect of this discussion with you all here today. We look forward to your knowledge, thoughts, and opinions, which will help us form the best possible guidance to industry.
Panels:
The most important topic in our initial tranche of guidance is also the subject of one of this morning’s panels: remote operations. Substantially all automated vehicle developers will have some remote interaction functionality. This may range from passive monitoring to communications with passengers and other road users. This may also include requests to remote human assistants to confirm it is appropriately contextualizing an environment in a binary “yes/no” capacity. Some developers provide such remote assistants with the ability to “nudge” a vehicle in to move in a particular direction, or to leave a trail of “breadcrumbs” to follow in an environment. On the far end of the spectrum, some developers may have a full set of manual driving controls at a remote location enabling a human to drive the vehicle. Each of these scenarios present different types and degrees of risk. We’ll hear more from our panelists on this topic, and we also want to hear from you this afternoon.
Today’s second panel involving safety metrics is also critically important. While NHTSA's immediate strategy to strike the balance between safety assurance and enabling innovation involves safety defect oversight and translation of existing FMVSS to better apply to new technology, we can’t stop there.
I am pleased to announce that NHTSA is working toward establishing minimum performance standards for ADS competency. This is not an easy task—no government in the world has been able to establish truly objective performance requirements that would meet the strict criteria for a FMVSS. But doing so is important to assure public safety and provide a stable and predictable environment for the technology to grow and develop right here in the United States.
Such standards require us to answer some challenging questions: what aspect of performance are we measuring? How do we measure it? What metrics do we use? What are the threshold levels of performance that provide safety—or “how safe is safe enough?” These questions may be simple, but they are not easy. As we go about our work in this space, I’ll be very interested in hearing from our panelists today, and from all of you in the future.
Modes
Before I wrap, I do want to say that this isn’t just a NHTSA effort. From the FAA to the FMCSA to the Federal Transit Administration’s work on automated buses for rural areas and supporting development of technologies to enable independence for people with disabilities, the entire Department of Transportation is engaged. We look forward to sharing the Department’s AV National Strategy as well.
In addition to convening today’s National AV Safety Forum, which we are all so excited about, we will also be holding discussions in the coming weeks with our co-regulators of AV technology at the state and local levels, and first responders.
As the industry moves from local pilots to commercial, multi-state deployments, this initiative is intended to better align federal safety oversight with state vehicle codes and local land-use policies. By bringing together regulators and the emergency responder community, NHTSA aims to build a shared framework that supports public trust and safety across jurisdictions while providing the regulatory certainty industry needs to scale operations nationwide. We want to ensure that as AVs cross state lines, the rules remain clear and safety remains the priority.
We are at the threshold of a transformational era in transportation. Thank you for being part of this journey. I can’t wait to see what we build together.
Thank you.