Transportation law is rapidly evolving with the advent of automation in the trucking sector. With 20 states allowing some form of automation, each with its unique set of rules and no overarching federal regulations, navigating this new terrain can be challenging for trucking companies. High-profile accidents involving automated vehicles (AVs) have garnered media attention, further complicating the issue. The Federal Motor Carrier Safety Administration (FMCSA) is moving toward enacting new regulations for the highest levels of automated driving systems (ADS). However, the FMCSA is not conducting further rulemaking for lower-level ADS involving a safety driver in a commercial motor vehicle (CMV). The present Federal Motor Carrier Safety Regulations (FMCSR) will continue to apply in those situations.
A review and assessment of the laws adopted by New York, Pennsylvania, Arizona, and Texas provide a spectrum of state-level oversight, moving toward full adoption of ADS rapidly in the continued absence of federal regulation of CMVs using ADS. We assess the risks of CMV ADS and the requirements of these state regulatory schemes. We also propose best practices for CMV carriers and risk managers seeking to reduce risk and prepare for the next era of litigation involving full and driver-assisted ADS as CMV carriers seek to reap the benefits of ADS, including probable reductions in equipment downtime and driver shortages.
ADS operates the vehicle with sensors, cameras, LiDAR (light detection and ranging), Global Positioning System (GPS) navigation, and mapping software to control driving actions. SAE International (formerly the Society of Automotive Engineers) has established a range of complexity for ADS from Level 1 to Level 5 (SAE J3016); this scheme has been adopted by the FMCSA. Level 1, the lowest level, includes cruise control, lane-departure warnings, and automatic emergency braking. Level 2 involves vehicle control of steering, acceleration, and braking while the driver remains engaged and ready to take over the wheel. Level 3 is close to full automation but in limited circumstances, such as on the highway, and requires a driver for local travel or to take over when necessary due to road conditions. Level 4 involves full automation but requires the vehicle to remain in some geographic regions or roadways, avoiding adverse weather conditions. At Level 5, the vehicle can operate autonomously in any circumstance and route.
Levels 1 through 4 ADS are currently being deployed in select areas by CMV carriers. Levels 1 and 2 ADS are more prevalent now, with CMV carriers using adaptive cruise control and lane-departure warnings to aid drivers with their routes. Companies are testing Levels 3 and 4 ADS in specific geographic areas or on particular routes. Level 3 is the highest level of ADS before there is no driver involvement; it involves extensive control of the CMV by the ADS. In contrast, Levels 4 and 5 ADS do not require a driver, which is a particular focus of current rulemaking.
Levels 4 and 5 ADS are being researched and developed in selected geographic areas. In 2019, the autonomous trucking company TuSimple began testing a Level 4 ADS on a 1,200-mile route between Phoenix, Arizona, and Dallas, Texas. The company used a combination of cameras, radar, and LiDAR sensors to enable its trucks to navigate highways and make deliveries without human intervention. The CMVs still contained safety drivers and engineers to ensure the ADS operated safely. Notably, the FMCSA got involved when a viral video showed a TuSimple vehicle in Arizona suddenly jutted to the left on a highway, striking a concrete median (no injuries resulted). The safety driver tried to counter-steer the truck following a computer-generated command several minutes earlier. The FMCSA closed the investigation without penalties, and testing of the Level 4 ADS continues, with improvements based on the event, according to TuSimple. Still, as reported by the Wall Street Journal, TuSimple’s safety drivers “have flagged concerns about failures in a mechanism that didn’t always enable them to shut off the self-driving system by turning the steering wheel” (Kate O’Keeffe & Heather Somerville, Self-Driving Truck Accident Draws Attention to Safety at TuSimple, Wall St. J. (Aug. 1, 2022)).
Waymo, a subsidiary of Google, has been testing ADS Level 4 since 2017. On May 5, 2022, a Waymo Via truck with a safety driver was pushed off the roadway by a non-ADS CMV in a hit-and-run accident outside Dallas, Texas. Waymo’s ADS and safety driver were determined not to be at fault. However, such incidents are red flags for the fledgling CMV ADS industry and raise questions regarding risk management, best practices, and the future of CMV ADS litigation.
Risks inherent in CMV ADS include sensor limitations (poor weather degrading LiDAR or radar performance), sensor fusion challenges (errors caused by the integration of different ADS), inadequate mapping (incorrect and outdated maps resulting in improper navigation of the road), software bugs (ADS failure caused by faults in the software code), hardware failure (malfunctions in sensors, actuators, or onboard computers), cybersecurity threats (hacking and malware to control or disrupt the vehicle), and edge cases (unusual or unpredictable situations, such as construction or unique traffic patterns). As the industry grows, these risks must be addressed.
The 2018 Uber crash in Tempe, Arizona, the first-ever pedestrian fatality involving a self-driving car, promptly led to Uber’s ADS program being revoked for testing in Arizona. The accident inflamed a sentiment that testing ADS on active city streets could be too dangerous to allow. Arizona’s state government had sought to attract technology companies, garnering investment and employment opportunities, by keeping regulation at a minimum. Soon, a video from the Uber vehicle was released, showing the driver looking down and not at the road. In total, the driver had traveled 3.67 miles while not looking at the road. At 9:59 pm, six seconds pre-impact, the ADS in the Uber identified the pedestrian while the vehicle traveled 43 mph. However, due to limited lighting, it classified the pedestrian as an unknown object, not a person, then as a bicycle. However, the ADS could not determine which direction the pedestrian was headed. At 1.3 seconds pre-impact, the ADS decided it needed to apply the brakes, but Uber had removed the ability for the vehicle to brake on its own.
The Uber accident is exemplary of an ADS-involved collision. A settlement resolved the pending litigation for this accident. Therefore, we do not know the basis of the alleged liability in this event. However, the facts alone show a potent mix of risks. An accident such as this would most likely involve allegations of traditional negligence on the driver for not paying attention, negligence through vicarious liability for that driver, negligence for disabling the brakes on the ADS, and product liability (sensor limitations due to low lighting).
The released police body camera footage showed further issues with ADS adoption. Specifically, the responding law enforcement officers appeared ill-equipped to investigate this type of accident, which was a new phenomenon. They hesitantly read Miranda rights to the driver while attempting to determine the cause of the accident. Was it the person or the machine? This event led Uber to shut down Arizona operations, laying off 300 employees (Matt McFarland, Uber Shuts Down Self-Driving Operations in Arizona, CNN (May 23, 2018)).
As noted, states such as Arizona and Texas have positioned themselves as hubs for autonomous vehicle testing and deployments, partly by creating regulatory landscapes that are easy for new companies to navigate. Nuro, a company making self-driving delivery vehicles partnering with major retailers such as CVS and Kroger, launched its service in Houston, partly because of the low regulatory hurdles in Texas. “Among other things, there is explicit authorization for AV operation, and the path to commercial deployment that not all states have,” said Katie Stevens, head of state and local policy at Nuro, based in Mountain View, California (Skip Descant, Regulatory Landscapes Determine Paths for AV Futures, Governing (July 23, 2021)).
Texas is at the open end of the spectrum regarding ADS adoption. Under Texas law, “An automated motor vehicle may operate in this state with the automated driving system engaged, regardless of whether a human operator is physically present in the vehicle” (Tex. Transp. Code, subch. J, sec. A545.454). The ADS is “considered to be licensed to operate the vehicle” (id.). A safety driver is not required if an ADS is engaged in the vehicle. The insurance requirements for ADS-equipped vehicles are the same as with any other vehicle.
Arizona is similarly open to ADS adoption. Under Arizona Revised Statute § 28-9702, autonomous vehicles must comply with all applicable federal and state laws, with or without a human driver. A person may operate an autonomous vehicle with the ADS engaged on public roads in Arizona with a licensed human driver who can resume part or all the dynamic driving tasks or respond to a request to intervene. Fully autonomous vehicles with no safety driver, ADS Levels 4 and 5, may operate on public roads if a law enforcement interaction plan and a written statement acknowledging specific requirements are submitted to the Arizona Department of Transportation with prior approval. The ADS is considered the driver or operator of the AV when engaged, and a licensed human driver is not needed when the CMV carrier complies with the interaction plan and written statement. The implications of an ADS being considered a “driver” is that the ADS must comply with the rules of the road, with the CMV carrier or AV vehicle owner responsible for any violations of those rules.
In November 2022, Pennsylvania passed new legislation that explicitly allows the testing or deployment of autonomous vehicles without a human operator. Pennsylvania law had previously prohibited vehicle operation on state roadways without a human driver inside the vehicle. The law also allows “platooning,” with a human driver in a vehicle leading other full ADS vehicles trailing the first on a route (Keith Goble, New Pennsylvania Law Permits Platooning, Land Line (Nov. 8, 2022)).
Meanwhile, New York State is currently at the restrictive end of the spectrum concerning adopting ADS. There is no regulatory framework for the commercial deployment of autonomous vehicles in New York, and ADS Level 4 is not allowed due to concerns about public safety. However, AVs with ADS Levels 1 through 3 are permitted for testing with a safety driver. To operate a semi-autonomous vehicle, a company must obtain a permit from the New York State Department of Motor Vehicles and comply with specific testing requirements. This limited AV testing program was to expire on April 1, 2023. New York Senate Bill S8468 was introduced in the 2021–2022 Legislative Session and updated on March 4, 2022, but has not passed the Transportation Committee. It would allow AV operation on public roads. The ADS would be considered the vehicle’s driver. The bill would set certain insurance requirements specific to vehicles with ADS, and the owner of the AV would be responsible for reporting accidents.
Regarding regulation on a national level, the FMCSA issued an advanced notice of proposed rulemaking in 2019 to obtain comments on the costs and benefits of future formal rulemaking for ADS Levels 4 and 5 automation. Then, on February 1, 2023, the FMCSA issued a supplement to the advanced notice of proposed rulemaking, titled Safe Integration of Automated Driving Systems (ADS)-Equipped Commercial Motor Vehicles (CMVs). The FMCSA is asserting rulemaking authority under the Motor Carrier Act of 1935 (49 U.S.C. § 31502), Motor Carrier Safety Act of 1984 (49 U.S.C. ch. 311, subch. 3), and Commercial Motor Vehicle Safety Act of 1986 (49 U.S.C. ch. 313).
The FMCSA’s primary focus is on ADS Levels 4 and 5. As to Levels 1 through 3, the FMCSA does not believe there is a need for new regulations because a licensed human CMV driver must always be seated behind the wheel to perform or be ready to take over dynamic driving tasks. As such, the FMCSR will continue to apply to those vehicles.
The proposed rulemaking outlines a plan for the NHTSA to establish guidelines for manufacturers of automated vehicle equipment and the FMCSA to focus on a uniform regulatory framework for automation in interstate commerce for ADS Levels 4 and 5 while keeping consistent rules for Levels 1 through 3. The FMCSA included the following questions (among others) for public comment:
The rulemaking process is ongoing and, given the breakneck speed of past development of ADS, may rapidly advance to answer the above questions as responses from industry and consumer advocates propose the path for future regulatory oversight.
Given this patchwork of state-level regulation and the ongoing and developing nature of federal oversight, CMV carriers and risk management professionals must remain diligent in navigating this new area of risk and potential litigation.
Litigation involving an ADS vehicle will broach both traditional negligence and product liability. Recommendations include monitoring ADS systems, with detailed documentation of testing and maintenance of the sensors, LiDAR, etc., used by the ADS to show compliance with best practices. For Levels 1 through 3 ADS, driver training programs will need to be updated to account for the ADS specifications and the required inputs by the driver. Compliance with the current FMCSR for commercial drivers is also necessary. Based on the recent supplemental advanced notice of rulemaking by the FMCSA, the logbook requirements and pre-trip inspections will still be in effect for Levels 1 through 3 ADS. For Levels 4 and 5 ADS, further testing of these systems is necessary before full adoption. CMV carriers should work closely with the manufacturer of the ADS system to monitor self-assessment of the system, safety procedures, and ongoing communication in case of issues with the system. Risk management should ensure proper coverage or indemnification by the ADS developer based on the ADS used. Should an accident occur, there must be rapid response by a team prepared to retain all ADS data, drive camera footage, and, if there is a safety driver in the vehicle, records for compliance with the FMCSR.
Adopting automation in the trucking industry presents unique challenges and opportunities for trucking companies. As transportation laws continue to evolve, companies must navigate a complex legal landscape that varies from state to state. By understanding the different regulatory schemes in place and investing in safety measures, companies can manage risk while reaping the benefits of Levels 1 through 4 automation. The potential reductions in equipment downtime and driver shortages make the effort worthwhile.
