While adaptive cruise control may seem very different from a fully autonomous vehicle — and in terms of technology, the two are on top of that. same spectrum. And even in the most advanced semi-autonomous technologies on the road today, humans are still required to be ready to drive the vehicle; that is, even if the machine has the baton most of the time, the human must be ready to grab it immediately when the machine doesn’t know what to do.
What does the handover model mean for truckers? In theory, trucks should handle most of the driving work in good conditions, and human truck drivers should take over in situations when machines fail—such as in construction zones or falls. crowded or when visibility is poor. Theoretically, when machines are in charge, truck drivers can be “not chained off the wheel.” and liberate for other missions.
This vision is similar to convert on the role of bank tellers after the advent of ATMs: Machines take on mundane routines, freeing people up for more interesting or skill-matched activities. But it leaves big questions about how truckers will be paid for time spent sitting in their vehicles while trucks drive themselves — after all, if trucking companies are still pay large labor costs, are self-propelled trucks worth the investment? necessarily deal with the issues surrounding overwork and fatigue.
There is another even more fundamental problem. The baton-passing technique is extremely difficult—perhaps confusing—very difficult to do smoothly in situations like driving. Remember that a machine transfers responsibility to humans in the situations it finds most difficult: when conditions are abnormal, when there is something in the environment that it is not equipped to deal with, when there is an emergency. machinery breakdown or emergency. Those situations are most likely safety-critical. One review of the academic literature found “a lot of evidence” that automating some aspects of driving has resulted in a “higher (almost) rate of collisions in critical events compared to manual driving… Basically, if the automation crashes unexpectedly with very little time for humans to respond, then almost all drivers crash.”
This problem is serious because the length of time the baton is transmitted is very small: Due to the nature of driving, humans can have a very short time—maybe just a fraction of a second—which understands machine intervention requirements, environmental assessment, and vehicle control. This tiny window of time is why human drivers in semi-autonomous cars are warned that they must remain vigilant the entire time the car is in motion. Despite the image of people relaxing, napping, texting, eating, and liberated from the demands of driving, this image is clearly unrealistic due to the need for fast, safety-critical transfers. at the current level of automation.
Audio and visual alarms can help people know when a handover is about to come, but the need for immediate control means that people still have to pay constant attention. However, NHTSA 2015 research found that in some cases it can take a full 17 seconds for people to regain control after a vehicle warns them to do so—far longer than it would take to avoid an accident.
