Autonomous truck maker Aurora Innovation Inc. is suing the Federal Motor Carrier Safety Administration after the agency denied its request for a five-year exemption from the placement of roadside warning devices in favor of cab-mounted warning beacons (CMWBs).
The filing, submitted Friday in the U.S. Court of Appeals for the District of Columbia Circuit, argues that the ruling is unlawful, “because it is arbitrary, capricious, [and] an abuse of discretion and otherwise not in accordance with law.”
This legal battle is the latest salvo in the ongoing debate between autonomous vehicle makers and the regulatory agencies that oversee them. The exemption was denied in late December. Aurora claims in the court filing that for nearly two years after receiving the company’s application, the FMCSA did not ask any additional questions even after both Aurora and Waymo submitted research confirming the benefits of CMWBs.
It’s all about equivalency for FMCSA
The crux of the legal issue is whether the CMWBs “achieve a level of safety that is equivalent to, or greater than, the level of safety that would be achieved absent the exemption.”
The current regulations say that if a driver of a commercial motor vehicle (CMV) is stopped on the shoulder of a road for any reason other than a necessary traffic stop the driver must activate hazard warning signal flashers and place required warning devices as soon as possible but within no more than 10 minutes at specified locations behind and in front of the CMV.
There are also regulations about placement of warning devices based on factors such as light levels and whether the vehicle is obstructed from view – for example, within 500 feet of a curve or the crest of a hill. For autonomous truck makers, this poses a challenge, since no driver is available to comply with the regulations. That issue prompted the exemption request and research studies.
The two separate studies submitted, one by the Virginia Tech Transportation Institute sponsored by Waymo, and the second by Aurora, looked at the responses of several thousand motorists across various lighting conditions and interstate roadway geometrics. “Both studies demonstrated that CMWBs are equally or more effective in enabling road users to detect, recognize, and react to the hazard presented by a commercial motor vehicle parked on a roadway as compared to human-placed warning devices,” Aurora’s filing states.
In layman’s terms, flashing lights appeared to be just as good as the warning devices.
But the data was not enough to sway FMCSA. The agency noted that the studies conducted by the Virginia Tech Transportation Institute and Aurora involved limited participant numbers and confined testing scenarios, which may not sufficiently demonstrate the equivalency or superiority of CMWBs over traditional warning devices in diverse real-world conditions.
The denial filed in the Federal Register gave more details on the research. The Aurora study was conducted in Texas on Interstate 45 between Exits 258 and 249, and the report did not describe the weather conditions. The Waymo-sponsored Virginia Tech study was conducted on a closed-circuit test track in daylight and nighttime conditions. Vehicles encountered a stopped CMV at three locations, including on the shoulder, in the lane ahead and on the right shoulder after a curve.
The FMCSA saw another concern, adding, “For example, as the Waymo study acknowledged, the unfamiliarity of the warning beacons may have contributed to motorist behavior.”
Interestingly, while the applicants in that study stated that the beacons were “easier to see” or “preferred” by drivers, that was not what they preferred. The FMCSA added, “The study also showed that for respondents who answered which device was better for signaling a stopped truck, 8 out of 10 selected warning triangles over beacons.”
The denial was equally scathing of the Aurora study, stating, “Aurora’s study notes that some motorist responses may have occurred wholly outside the sensor range.”
Another issue, according to the FMCSA, is that both studies fail to fully support whether the motorists’ behavior was causally based on seeing warning beacons or on seeing a stopped CMV. Think of it like the chicken and the egg paradox, except the chicken is 13.5 feet tall and weighs 80,000 pounds, while the egg is a lightweight, 17-inch-tall highway triangle. For the FMCSA, it’s paramount to know which one drivers spotted first.
Did agencies test original regulations in 1972?
For Aurora, the FMCSA’s scathing assessment raises the question of whether such due diligence was required for the warning triangles in the first place.
Aurora argues that the data it provided “far exceeds what FMCSA has admitted about the agency’s existing requirements: FMCSA has ‘never conducted experimental research on the impact of using’ the currently-mandated human-placed warning devices, and it is ‘historically unresolved’ whether ‘the use of such [driver-placed] devices improves traffic.’”
Aurora added that no actual data was presented to support the safety benefit of specific human-placed warning devices when the original regulations were hatched in 1972. Moreover, placing the devices may be hazardous to the driver, it argued.
Aurora wrote: “Under current regulations, when a CMV is stopped on a highway, the driver must, within 10 minutes, exit the truck, walk onto a roadway that may have traffic and other conditions, and place the required warning devices in three locations around the CMV, exposing the driver to the significant dangers presented by roadway traffic.”
The petition concludes that the denial on Dec. 26 is at odds with the law and the FMCSA’s own public commitment to support safety and innovation.
FMCSA launches its own study
Following its denial of an exemption for Aurora, the FMCSA initiated an experimental study to determine whether CMWBs significantly influence “crash-relevant aspects of human performance” around stopped CMVs.
The FMCSA is seeking 256 drivers to collect data for the study and test the warning devices, “in a controlled experiment at a closed-course, state-of-the-art driving facility that will allow the most comprehensive examination of the effects of warning devices to date.”
While the FMCSA makes the rules, it’s the National Highway Traffic Safety Administration that dictates the performance and design specifications for warning devices. This can include, in the case of a warning triangle, its reflectivity, color and luminance, among other specifications.
Further complicating the issue, involvement by the Federal Highway Administration in previous studies of automated driving systems has not added clarity, the study request states.
“Past attempts by the Federal Highway Administration (FHWA) and other researchers to answer those questions yielded generally inconclusive or inconsistent results, which possibly influenced NHTSA’s past decision not to pursue conducting its own research on the topic. FMCSA (previously under FHWA) itself has never conducted experimental research on the impact of using warning devices. As the only regulatory authority which still requires CMV operators to use warning devices, the responsibility to answer these questions finally and definitively is best charged to FMCSA.”
The notice adds that the last time these questions were explored was the 1980s, adding that “advanced research instruments unavailable or not in use at the time of all past research on this topic are now in common use and would permit far more sophisticated analyses of the effects of warning devices on driver behavior.”
While the FMCSA conducts its own research, autonomous truck makers plan to continue testing. Aurora, which plans to launch its driverless trucks in April, says the exemption denial will not prevent it from complying with existing regulations but raises a more important question about safety and innovation for America’s roads.
The post Autonomous trucking’s triangle-shaped problem goes to court appeared first on FreightWaves.
