By: Rachel McCleery, Executive Director, Coalition for Reimagined Mobility, and Amitai Bin-Nun, Director of Safety, Foretellix
Breakthroughs in physical AI are fast-tracking the evolution of autonomous driving. Look no further than the slew of headlines from the past couple of months. Waymo is rapidly expanding its national footprint, recently adding Dallas, Houston, San Antonio, and Orlando to the list of U.S. cities it operates in. Aurora’s driverless trucks can now travel 1,000 miles nonstop and recently completed a total of 250,000 miles of driverless travel. That’s the equivalent of roughly 10 trips around the Earth.
Coupled with this private-sector momentum is a new opportunity for the U.S. Department of Transportation (DOT) to take a major step forward in representing our country’s autonomous vehicle (AV) safety priorities on the global stage… As long as safety is broadly interpreted to include the preservation of national security interests for the U.S. and its allies.
The United Nations Economic Commission for Europe (UNECE)’s World Forum for Harmonization of Vehicle Regulations in January issued a draft rule requiring automated driving systems to perform at least as well as a competent human driver and for manufacturers to show clear evidence that their vehicles don’t pose unreasonable risks. The rule also establishes continuous safety monitoring with data recorders and fleet reporting, so real-world incidents and close calls are tracked. Much like our regulatory process in the U.S., this draft document is meant for organizations like SAFE to weigh in on and NHTSA and its foreign counterparts to eventually adopt.
With support from Foretellix, SAFE submitted comments to DOT on this draft rule endorsing international safety harmonization that ensures data access, in-service reporting, and safety-case documentation are protected from coerced disclosure, surveillance, or reverse engineering. In SAFE’s comments, protection across all relevant domains is addressed: data storage, supply chain security, intellectual property, and reporting and monitoring practices.
This is consistent with previous UNECE engagements. SAFE through Foretellix has participated in the UNECE’s World Forum for Harmonization of Vehicle Regulations as an accredited expert since 2019. In this and related forums, SAFE and its representatives drove the themes of enabling innovation while pursuing safer, more secure, and more efficient systems. Major accomplishments include advancing a new technical concept—coverage—as a key determinant of AV safety.
UNECE’s most recent action is timely as the auto industry grows increasingly more complex with the evolution of physical AI. Vehicles are becoming innovative tools with compute power needs and software stacks that have opaque supply chains. Their overall function remains the same: to safely move people and goods from one place to another. Technological innovation will only strengthen vehicle safety while also making the transportation ecosystem cheaper, faster, and more efficient. But without decisive and strategic action from the federal government, our country risks eroding competitiveness and exposing these innovative vehicle systems to cyber threats. This risk only increases as adoption and innovation outpace responsible federal oversight.
There’s already federal precedent to be aware of. A little over a year ago, the U.S. Department of Commerce banned connected-vehicle software and hardware tied to Chinese and Russian companies. SAFE’s written comments that the U.S. needs a clear and deliberate playbook for stronger cybersecurity standards, diversified and transparent supply chains, deeper cooperation with allies, and long-term investment in domestic connected-vehicle capabilities were adopted. DOT has the opportunity to finalize a rule consistent with the spirit of Commerce’s action.
The race to advance the adoption of autonomous vehicles is intrinsically tied to who deploys generative AI at scale, who governs it responsibly, and who sets the rules that others follow. Technical mandates are frameworks for data governance and national security, which are a type of safety standard. We urge DOT to ensure that no provision allows ease of data access to compromise the security interests of the US and its allies.
Foretellix is the leading enabler of safe autonomous vehicles. Its Physical AI toolchain uses real-world data curation, verification & validation (V&V), and synthetic data generation (SDG) to accelerate the training, validation and safety evaluation of autonomous vehicles, making AI safe for the physical world. With offices in the US, Europe, and Asia, Foretellix is driving the shift to intelligent and safe AI-powered autonomy.
Twitter
