Humans drive over ten trillion miles a year. The idea that using robots as drivers could reduce traffic deaths to zero is implausible. But are we really prepared for killer robo-drivers?
Robots drivers, we’re told, could reduce traffic deaths. That might be possible someday. But it is highly implausible that switching entirely to robot drivers will eliminate traffic deaths, which leads to the question: are we really prepared for killer robo-drivers?
Ask yourself this: How would you feel if it was your child, partner, parent or friend that was killed in one of the two Boeing 737 Max crashes? I’d feel angry, and the more I learn about the failings of Boeing’s Maneuvering Characteristics Augmentation System (MCAS) in the 737 Max, the angrier I feel. Reading the recent article on the subject in EE Times, one comment in particular stood out:
Ultimately, fuel savings, longer range and reduced operational costs appear to outweigh safety considerations.
Translation: Money beats safety.
The global 737 Max fleet was grounded for 20 months following the second fatal crash (in March 2019) while the role of MCAS was investigated. This gives us some major clues regarding the level of acceptance to software killing people. Amid all the hoopla and magical thinking surrounding the self-driving circus and the boasts of “saving lives,” little is said about where we are currently headed.
Are we really prepared for killer robo-drivers? Has anyone seriously thought through the legal implications of software killing other road users? How does public opinion of robo-drivers change as the body count rises? How will lawmakers and regulators respond to that change in public opinion? How is justice seen to be done when robo-drivers kill? We don’t know.
While the self-driving bandwagon keeps rolling on, both public opinion and consumer desire look to already be headed in another direction. In February the American Automobile Association (AAA) published a report entitled “Today’s Vehicle Technology Must Walk So Self-Driving Cars Can Run” with the main finding “Drivers say improving vehicle safety systems takes priority over developing self-driving cars.” The report goes on to say:
AAA’s annual automated vehicle survey finds that only 22% of people feel manufacturers should focus on developing self-driving vehicles. The majority of drivers (80%) say they want current vehicle safety systems, like automatic emergency braking and lane keeping assistance, to work better and more than half — 58% — said they want these systems in their next vehicle.
Which tells us decisively where public opinion in the U.S. stands. Consumers want human-driven cars with better safety systems, not self-driving vehicles with robo-drivers. But are automakers listening?
Hoopla beats safety
At the beginning of March, Honda announced what it claimed to be the world’s first Level 3 passenger car. If the goal was to generate spectacular press coverage, then Honda succeeded with articles here, here and here. As always, what matters isn’t the headline but the detail and I am left feeling decidedly underwhelmed. Let’s start by watching the dreamy promotional video and then go deeper.
Do you see what is missing? Traffic. Adverse weather. Construction zones. Road debris. Complex scenarios. The video doesn’t show any kind of everyday driving that I recognize. In fact, I yearn to drive myself on precisely these empty, straight roads with nothing but a full tank of gas and half a pack of cigarettes. Why pay big bucks for automation to do the fun, easy stuff?
Honda calls the system “SENSING Elite”, the main features of which seem to be automatic lane change and traffic jam assist. Audi, BMW, Mercedes-Benz and other automakers have offered traffic jam assist for several years, while GM’s Enhanced Super Cruise, introduced in the 2021 Cadillac Escalade, supports an automated lane change function as described in this video.
These systems are all marketed as Level 2 and it is not immediately apparent to me what the benefits of the Honda system are compared with those offered by other automakers. SENSING Elite looks like a convenience feature for highway driving that does little to increase safety and a lot to increase mode confusion.
As I have written previously, the critical challenge for safe deployment of Level 3 systems on public roads is navigating the “machine-to-human handover,” when the machine-driver issues a handover request to the human-driver. In research entitled “Transition to manual: driver behaviour when resuming control from a highly automated vehicle” (Merat, N. et al., 2014.), the authors observed:
This study also suggests that drivers require around 40 seconds to resume adequate and stable control of driving from automation, which might be considered a ‘comfortable transition time’ as stipulated by the NHTSA guidelines on Level 3 automation.
EE Times has written about the Level 3 puzzle and I look forward to reading further announcements from Honda explaining how it has resolved the handover issue. I am also looking for Honda to provide an unambiguous answer as to which party is legally liable and who is accountable for accidents or deaths that occur during the handover period.
The production run for the Honda Legend with the Level 3 system is limited to a paltry 100 units. Further, Honda is leasing not selling these vehicles and availability is restricted to Japan. Therefore, I expect the answers to the accountability and liability questions are buried deep in the detail of the lease contract.
In an interview, Yoichi Sugimoto from Honda R&D said Level 3 technology “is what will be needed in the future for sure,” and that it would eventually be distributed “hopefully in a decade or two” for one of Honda’s top-selling models.
Honda landed some big headlines with the SENSING Elite announcement, but it sounds to me little more than a marketing stunt tied to an R&D program, as opposed to anything which truly revolutionizes vehicle safety on public roads. My conclusion? At Honda, hoopla beats safety.
Safety comes first
The results from the AAA survey tell us that consumers want human-driven cars with better safety systems, not self-driving vehicles with robo-drivers. By prioritizing technological hoopla over safety, Honda appears not to be listening. Based on recent press activity, neither is Volkswagen, which has announced “Trinity,” a new platform that will be “technically ready for Level 4” — whatever that means — by the planned start of series production in 2026.
Tesla’s attempts to win the race to Level 5 have now backfired spectacularly. If automakers are going to compete for technological supremacy based on the levels of driving automation as specified in SAE International’s standard J3016,then it falls to bodies such as NCAP (New Car Assessment Program) and agencies such as NHTSA to continually advance safety standards in human-driven vehicles.
A recent article entitled “The US Invented Life-Saving Car Safety Ratings. Now They’re Useless,” explains in detail how NCAP helped prove that car-buyers care about safety. In it, the author makes a bold proposal to “not just overhaul NCAP’s tests, but to remove the program from the government entirely so it no longer has to adhere to this meandering rule-making process.” This suggestion has much merit.
Humans drive over ten trillion miles a year, resulting in more than 1.3 million traffic-related deaths globally. The idea that using robots as drivers could reduce traffic deaths to zero is statistically implausible, yet both Honda and Volkswagen are actively discussing robo-drivers before the accountability and liability issues have been resolved. Once more: Who is liable when robo-drivers kill?
Honda’s tagline is “The Power of Dreams,” but road safety standards are raised further by the mandatory installation of mundane safety technology (such as automatic emergency braking, lane keep assistance and driver monitoring) into human-driven vehicles than by dreamy videos promoting Level 3 systems such as Honda SENSING Elite.
Are we really prepared for killer robo-drivers? Just ask any of the families of those killed in the Boeing 737 Max crashes. On public roads, as in aviation, safety comes first.
Colin Barnden is principal analyst at Semicast Research and has over 25 years of experience as an industry analyst. He is considered a world expert on market trends for automotive vision-based driver monitoring systems (DMS). He holds a B.Eng. (Hons) in Electrical & Electronic Engineering from Aston University in England and has covered the automotive electronics market since 1999.