Automotive > In-Vehicle UX Blog

NHTSA: Unsafe at Any Speed

by Roger Lanctot | Dec 31, 2020

FCC Chairman Ajit Pai recently announced that he will step down from his post on January 20th, the day President-elect Joseph Biden will be sworn in. Pai will be remembered for voting against “Net neutrality,” but also for opening up additional Wi-Fi spectrum at the expense of the automotive industry.

The latter decision occurred weeks ago, near the very end of his term and following extensive research regarding potential interference from applications using unlicensed Wi-Fi spectrum impinging on 5.9GHz spectrum previously reserved for automotive safety applications. Pai made the move in the face of concerted auto and transportation industry objections – but the move came in the wake of 20 years of inactivity within that spectrum.

More importantly, Pai’s decision can be seen as a wake-up call to the automotive sector. After 20 years of testing, development, and standards-setting activity and the expenditure of an estimated $2.5B, the automotive industry has little to show in the way of on-the-road, in-vehicle deployments for its privileged access to spectrum in the 5.9GHz band.

Pai’s decision has forced NHTSA, the U.S. Department of Transportation, state-level DOTs, and vehicle manufacturers to shift their research focus away from wireless “talking car” technologies and towards sensor-based automated driving technologies intended to enable collision avoidance and automated driving today. The real challenge facing the agency – NHTSA – is to take on automated driving standards-setting and regulatory activity – which has been long-neglected.

As if on cue, the NHTSA published an 18-page Advance Notice of Proposed Rule Making on December 3 titled: “A Framework for Automated Driving System Safety.” This new NPRM, coming at the close of U.S. DOT Secretary Elaine Chao’s term, caps four years of inaction at NHTSA reflected in the fact that the agency was never assigned an Administrator by the Trump Administration or Chao.

NHTSA - "A Framework for Automated Driving System Safety" - https://www.govinfo.gov/content/pkg/FR-2020-12-03/pdf/2020-25930.pdf

The new ANPRM reads like a plea for help to make sense of all the conflicting forces being brought to bear on automated driving including everything from data access, to artificial intelligence, to identifying and addressing particular societal transportation challenges, to defining processes for certifying self-driving vehicles and providing exemptions for vehicles that do not meet current safety standards (think Nuro’s safety bots, which were recently approved).

The new NPRM concludes with 24 questions regarding which the agency is seeking input and this, above all, speaks to the lack of direction being experienced by the agency. NHTSA is literally begging for help to process and prioritize the requirements of automated driving so that it might maximize societal benefits while minimizing harms. 

The agency is asking via the NPRM whether it needs to go faster or slower or simply stay out of the way in order not to impede progress. The reality is that the agency knows only one speed – slow – and the NPRM is nothing less than an existential proposition that boils down to a single question: Is NHTSA equal to the task of regulating autonomous vehicle technology?

In contrast, the United Nations Economic Commission for Europe put out its own road map regarding Level 3-related driving standards and requirements with a focus on automatic lane keeping systems (ALKS) – as part of its own self-driving regulatory framework. The document spells out in detail a comprehensive vision for implementation and deployment of this technology.

UNECE - "Proposal for a new UN Regulation on uniform provisions concerning the approval of vehicles with regards to Automated Lane Keeping System" - https://undocs.org/ECE/TRANS/WP.29/2020/81

These two standards-setting agendas – UNECE vs. NHTSA – reveal one agency focusing on machine-assisted human-centric driving in the form of ALKS technology, while the NHTSA is seeking to grant exemptions to FMVSS (Federal Motor Vehicle Safety Standards) for self-driving cars operating without steering wheels. The existence of these two initiatives reveals that self-driving car regulation is inherently political. How you choose to regulate self-driving technology tells the world how you see the technology being adopted and in whose interest.

Regulators are facing very real battle lines over such fundamental questions as why autonomous vehicle technology is being pursued at all. The UNECE is placing its regulatory bets on self-driving vehicle technology as an enhancement to rather than a replacement of human driving. The UNECE guidance reflects the reality that fully automated driving is likely a decade off in the future, but that humans ought to benefit today from the advances in sensor and software systems capable of making driving safer.

NHTSA appears to be responding to pressure from commercial interests in the form of General Motors’ Cruise Automation and Alphabet’s Waymo, both of which are seeking to create automated-vehicle-based transportation networks. As a step in that process, both organizations are seeking approval to be exempted from including steering wheels in their self-driving vehicles.

Prioritizing steering wheel removal is a political decision. It is pushing expensive self-driving robotaxis to the front of the self-driving vehicle line, based on little more than lobbying muscle – especially given the presence of former senior NHTSA executives in senior regulatory roles at Waymo, Cruise, and Zoox. Zoox is the big kahuna of robotaxi regulatory capture with former (and last) NHTSA Administrator Mark Rosekind acting as Chief Safety Innovation Officer.

A greater concern for NHTSA might be the kind of vehicle configuration envisioned by both Cruise and emerging competitor, Zoox, both of which have proposed and prototyped in-vehicle seating configurations with facing passengers – creating an entirely new safety challenge as-yet unaddressed in the regulatory canon. One has to ask what is considered a safe speed for a pilotless vehicle with passengers seated facing one another – NHTSA would normally expend millions of dollars and years of research to answer that question alone.

It is also possible that some at NHTSA envision robotaxis serving the needs of underserved or disadvantaged populations – improving transportation access. Given the poor real-world performance of operators such as Uber and Lyft in fulfilling these broader societal aims, it is fair to consider such ambitious expectations for robotaxi deployments to be naïve, at best.

Meanwhile, UNECE has laid out an elaborate regulatory framework and implementation guidelines, supported by extensive research, for the widespread adoption of automatic lane keeping system technology. The UNECE regulation:

Defines safety requirements for:

  • Emergency Maneuvers, in case of an imminent collision;
  • Transition Demand, when the System asks the driver to take back control;
  • Minimum Risk Maneuvers - when the driver does not respond to a transition demand, in all situations the system shall minimize risks to safety of the vehicle occupants and other road users.

And:

ALKS activation criteria:

  • The driver is in the driver seat with safety belt fastened;
  • The driver is available to take over control of the driving task;
  • No failure affecting the safe operation or some functionalities of the system is detected;
  • DSSAD (data storage system for automated driving) is operational;
  • Positive confirmation of system self-check; and
  • The vehicle is on roads where pedestrians and cyclists are prohibited and which, by design, are equipped with a physical separation that divides the traffic moving in opposite directions;
  • The environmental and infrastructural conditions allow the operation;

Driver Availability Recognition System:

  • Driver presence;
  • Driver availability;
  • Actions taken when driver is deemed unavailable.
  • Criteria for deeming driver availability:
  • The driver deemed to be unavailable unless at least two availability criteria (e.g. input to driver-exclusive vehicle control, eye blinking, eye closure, conscious head or body movement) have individually determined that the driver is available over the last 30 seconds;
  • Actions taken when driver is deemed unavailable.

Data Storage System for Automated Driving (DSSAD)  

The system will record the following events:

  • Activation of the system;
  • Deactivation of the system (e.g. override on the steering wheel);
  • Transition Demand by the system (e.g. planned, unplanned etc.);
  • Reduction or suppression of driver input;
  • Emergency Maneuver;
  • Involved in a detected collision;
  • Minimum Risk Maneuver engagement by the system;
  • Failures.

In this way, UNECE is taking on the task of regulating an active safety system designed to avoid collisions. The regulation is intended to address the demands of so-called SAE (Society of Automotive Engineers) Level 3 driving in which the driver is expected to supervise automated driving tasks while simultaneously being assisted by the self-driving system.

The UNECE Regulation goes so far as to require that “on-board displays used by the driver for activities other than driving when the ALKS is activated shall be automatically suspended as soon as the system issues a transition demand, for instance in advance of the end of an authorized road section. The Regulation also lays down requirements on how the driving task shall be safely handed back from the ALKS to the driver, including the capability for the vehicle to come to a stop in case the driver does not reply appropriately,” according to a UNECE newsletter on the topic.

The Regulation was drafted by experts in Japan and Germany and is expected to see widespread adoption across Europe and Canada when it enters into force in January. The Regulation was agreed to by 60 countries within the UNECE’s World Forum for Harmonization of Vehicle Regulations.

The UNECE Regulation is critical because most car makers are introducing or have introduced so-called Level 2+ systems (Tesla Motors Autopilot, General Motors Super Cruise, etc.) that typically combine lane keeping systems with adaptive cruise control to deliver a hands-off driving experience that is one step away from a Level 3 proposition. The UNECE Regulation lays out the requirements for the introduction and configuration of these systems including the in-vehicle user interfaces and driver monitoring requirement.

NHTSA, meanwhile, is asleep at the wheel – and that somnolence is reflected in the agency’s reaction to the 5.9GHz spectrum re-allocation action by the FCC.

Representatives of the NHTSA told the Automotive News Shift podcast in December that the agency is still weighing its options in the wake of the FCC spectrum decision. Finch Fulton, U.S. deputy assistant secretary for transportation policy, noted on the podcast: “(With the FCC removing half of the spectrum) the unlicensed Wi-Fi will bleed into the (safety) spectrum. If you need these signals for safety you can’t have wi-fi bleeding into it. By taking this step it ruins the value of the spectrum which then gives freedom to give away the rest of it if the technology can’t work.

“It’s an existential threat to the future of connected vehicle technologies. It has widespread safety and efficiency impacts. It’s not over yet. I don’t know the status of any lawsuits in the industry. I’ve heard some things. It’s concerning. We haven’t gotten our way and I think it is a fight that will continue on into the future.” Please, no.

Automotive News "Shift" podcast - "USDOT Sketches Road Rules Ahead for Self-Driving Vehicles" - https://www.autonews.com/shift-podcast-about-mobility/usdot-sketches-road-rules-ahead-self-driving-vehicles-episode-74

Suffice it to say that senior NHTSA leadership is still clinging to its vision of continuing to deploy dedicated short range communication (DSRC) Wi-Fi technology for inter-vehicle and vehicle-to-infrastructure communication. NHTSA is not picking up the wake-up call from the FCC’s Pai.

The message from the FCC is that NHTSA should shift its focus to cellular-based inter-vehicle communications along the lines of 5G and C-V2X and should make this work within the confines of the re-allocated 5.9GHz spectrum. Case closed.

Even more important, though, is the message that more immediate, life-saving, technology standardization activities are within reach and within the purview of the NHTSA and worthy of greater focus today. V2X communications are a longer term project and offer little hope of substantial life-saving outcomes in the short-term.

Just look to the UNECE. The standards-setting and regulatory activity of the UNECE are derived from an existing framework – something the NHTSA is seeking help in creating for itself. NHTSA might save some time and effort by aligning with the UNECE in the interest of saving lives today facilitating the adoption and deployment of collision avoidance technology in cars that are being sold today. 

The UNECE has put forward a compelling vision reflecting the regulatory alignment of 60 countries preparing for imminent adoption. This is no time for the leading U.S. regulatory body to go it alone when it comes to saving lives. We have the technology to save lives today in the U.S. We just need some regulatory guidance. We don’t need a prolonged battle over Wi-Fi spectrum.

 
 
Previous Post: No iCar for Me, Thank You | Next Post: 2021: New Priorities for Map Makers

Let's talk

Now you know a little about us, get in touch and tell us what your business problem is.
Name:
Email:
Telephone:
Country:
Inquiry / Message: