Automated Driving Meets Regulation NHTSA and the Next 50 Years

May 1, 2016

Daniel P. Malone and John F. Creamer

HTSA Administrator Mark Rosekind recently spoke on the prospects for reaching zero deaths on America’s roads. “Conceptually, the challenge is extremely straightforward. If we keep doing what we are doing, we are never going to get there,” he said. “We’ve got to keep doing what we’ve been doing — it’s been successful — but we’ve got to do more.”

Dr. Rosekind and others can talk about zero road deaths because the automotive industry has developed, without any legal requirement to do so, automated systems that compensate for driver error. However, these systems cannot be effectively regulated under the current U.S. safety regime. In fact, NHTSA and the FMVSS (Federal Motor Vehicle Safety Standards) were established precisely because the technology to avoid accidents did not exist in the 1960s and 1970s.

In order for NHTSA to “do more,” however, Congress will need to enable NHTSA to “do differently.”

Recall system rooted in 1960s

Fifty years ago, Congress set up the U.S. regulatory system to force automakers to design safety measures into their vehicles, in large part because car buyers at the time broadly rejected the added cost of seat belts, interior padding, and other safety features. Because this approach is fundamentally coercive, the courts, Congress, and the Executive Branch have placed constraints on NHTSA rulemaking such that any change in requirements can take a decade of painstaking research to justify.

But NHTSA no longer has the luxury of time, given the rapid development of new automated-driving technologies and safety systems. Unable to go through its rigid, step-by-step administrative process, NHTSA has resorted to going around it. During the past six months, NHTSA has pursued voluntary agreements with automakers on automatic emergency braking, cybersecurity, recall system improvements, and an upgrade of the consumer-oriented New Car Assessment Program (NCAP). There’s also a $4 billion budget request to promote automated vehicles, including a promise to develop “guidance” on the safe deployment of these technologies in cooperation with industry.

Congress also set up the safety-defect reporting and recall system based upon 1960s-era notions of consumer protection. NHTSA receives roughly 330 calls per day from vehicle owners, but these calls rarely provide the technical detail necessary to isolate a potential problem (as noted in the Inspector General’s June 2015 report on NHTSA’s Office of Defect Investigations). Research at the University of Iowa and consumer surveys have found that most drivers are unsure about how advanced vehicle systems work or why their vehicles behave the way they do.

Faced with the inadequacy of this “consumer-complaints bureau” approach, Congress has increasingly placed the burden on automakers to report any suspicions of a defect, with severe penalties for non-compliance. As a result, NHTSA receives some 6000 quarterly early-warning reports from manufacturers each year while the number of recalls has risen to levels that defy public understanding. That’s even after excluding the high-profile cases that have sent recall volumes in excess of 50 million vehicles for two years running.

Despite all these efforts — or arguably because of them — this system requires years of reporting and accident data to isolate a potential problem. It takes several years more (if ever) to find a root cause, and then fails consistently to see defective vehicles repaired. According to the consumer website Carfax, more than 47 million cars are running on American roads with an unfixed recall issue.

Standards-based rulemaking needed

Secretary of Transportation Anthony Foxx declared that automated vehicles will become widespread within 10 years. However, their complex electronic, software-driven systems are open to risks and technical faults that electrical and mechanical systems are not. Beyond well-known concerns over malicious hacking, computer platforms are routinely and rapidly personalized by their owners through choices of features to activate, downloads of third-party applications, updates (or the failure to update), system upgrades (which can be incompatible with older hardware), and other decisions. Motor-vehicle software will increasingly be serviced through over-the-air (OTA) updates, raising both opportunities and risks.

In effect, we can expect that vehicles in the future will be subject to in-use safety failures resulting from circumstances unique to each vehicle where identifying root causes may be impossible.

NHTSA has a critical role to play in facilitating the smooth and safe deployment of the new safety technologies. The agency’s leadership can guide the development of uniform, clear, and legally enforceable nationwide standards. NHTSA leadership can guide reform of the safety-defects system to better protect the public welfare. But only Congress can enable NHTSA to fulfill its mission.

The challenges and opportunities on the road to “zero deaths” demand a coherent, national (and even international) response. Solutions must be found quickly, but not hastily, through greater transparency and collaboration among stakeholders. We are entering what is arguably the final frontier in road safety, one in which traffic and vehicle safety merge. The U.S. regulatory system needs to move with it and Congress needs to make this happen.

Rulemaking in this new era needs to more closely resemble standards-setting within SAE International. Rather than focusing on mandatory vehicle performance, a new level of federal automotive safety technical standard should focus on ensuring that systems installed on vehicles are safe in the hands of the average driver. These standards should set uniform minimum requirements for safety, privacy, security and other aspects in the public interest.

Unlike the FMVSS, however, these technical standards would not mandate vehicle performance. Rather, they would set performance requirements for automated systems. A vehicle would be required to meet these requirements only if fitted with the given system; the FMVSS still could be used to mandate their use, but these federal automotive safety technical standards would leave use of the systems up to each automaker.

As such, NHTSA should be enabled to establish the standards more quickly, without the extensive cost-benefit justifications required of FMVSS. For this to happen, NHTSA needs to collaborate with stake-holders in road safety and NHTSA is absolutely correct to be reaching out to automakers for support in developing standards. However, within a regulatory context, collaboration needs to go beyond agreements with specific companies.

NHTSA standards should be transparently based upon consensus among automakers, the supplier industry (including independent aftermarket manufacturers), the vehicle service and repair industry, and state traffic authorities. Participation should be open to industry standards-setting bodies such as SAE International and research institutions such as the University of Michigan Transportation Research Institute (UMTRI), the Center for Automotive Research, and the Virginia Tech Transportation Institute, as well as independent testing groups such as the Insurance Institute for Highway Safety (IIHS). In order for standards-setting to be credible, all stakeholders, including safety advocates, need to be part of the process.

Robust OBD required

The credibility of these standards becomes especially important given the nature of automated driving systems. Unlike most safety technologies, automated driving systems interact with the driver and with other vehicles and road users. Automation fundamentally changes transportation and such change is never without risk. While these systems will prevent millions of accidents and injuries and save thousands of lives, they will not be perfect from the outset.

Even with standards such as SAE’s ASIL (Automotive Safety Integrity Level, a risk-classification protocol), it is a virtual certainty that automated systems will, at some point, encounter complex conditions and fail to perform as desired. In such cases, the NHTSA technical standards should enable differentiation between unanticipated outcomes and liability under recognized legal theories.

The standards can play an important role in promoting both high levels of performance and greater legal certainty to facilitate the safety transformation.

Given the technical complexity of automated systems, the safety-defects system cannot rely mainly upon accident reporting, consumer complaints, and other sources that do not provide consistent technical information on vehicle behavior, especially prior to death or injury. Automated systems will require robust on-board diagnostics (OBD) and data recorders. These technologies open avenues for monitoring vehicle behavior toward identifying anomalies or unforeseen effects on traffic patterns or driver behavior that can be used to continuously improve the system standards.

At the same time, car owners are fundamentally responsible for ensuring the proper maintenance of their vehicles. In a future of connected and interdependent vehicles, the public cannot afford to have millions of cars with malfunctioning systems on the roads. Congress should enable NHTSA to work with state authorities and the automotive manufacturing and service industries to devise a uniform national system for ensuring that critical vehicle systems remain in proper working order — and any problems remedied before they evolve into dangerous malfunctions.

All of this presupposes an unprecedented level of cooperation among stakeholders, but this collaboration is the future. A world of connected vehicles requires a community of connected stakeholders to ensure safety.

Since 1946, U.S. traffic fatalities have never fallen to less than 30,000 per year. This sad reality can change because automated driving technologies address the root cause of 94% of all accidents. At the same time, unlike any safety system previously introduced into motor vehicles, automated technologies fundamentally change the nature of driving.

We cannot meet the challenge under a system designed in the 1960s based on assumptions from the 1950s and technologies from the 1970s. As it did 50 years ago, Congress must act boldly to refocus road safety in America for the next 50 years. The new safety era requires a new approach if the nation is to realize the potential of these advanced technologies and ensure U.S. leadership in this economic and social transformation.

This article is based on a longer technical paper (2016-01-7000) by the authors and published by SAE International in April. It can be accessed here: http://papers.sae.org/2016-01-7000/ .

Daniel P. Malone, an attorney with Butzel Long, specializes in automotive product safety-related litigation and regulatory matters. He publishes articles and presents on legal and automotive-related issues frequently. Mr. Malone can be reached at This email address is being protected from spambots. You need JavaScript enabled to view it..

John F. Creamer is Managing Director of GlobalAutoRegs.com, an advisory service on international automotive regulations. He is also a partner in The Potomac Alliance, a global consortium of regulatory affairs experts. Mr. Creamer can be reached at This email address is being protected from spambots. You need JavaScript enabled to view it..

---