I just received City Councilor Jake Auchincloss’s always informative email newsletter about what’s going on in the city. Here’s the most interesting item in this month’s issue….
Newton will join 13 other Greater Boston cities and towns, including Boston, in permitting companies developing autonomous vehicles (AVs) to test on public roads. Coordinated by the governor and the Metropolitan Area Planning Council, the 14 jurisdictions will share one application portal. Each jurisdiction retains the right to exclude any company or use-case, even those approved under the application.
The AVs will have a human behind the steering wheel who is ready to assume control if necessary. They will not be carrying passengers or otherwise engaging in activities outside of testing, and they must prove competence through progressively more complex driving scenarios. The application process is being finalized, now, and I expect the first AVs from local companies like Optimus Ride and nuTonomy will be on Newton’s roads in 2018.
The benefits to Newton are three-fold. First, the city can shape protocols before they become codified. For example, the Newton Fire Department recommended that fire fighters from all jurisdictions receive training in how to safely extract persons from damaged AVs, which may require cutting through hardware that is different than on regular vehicles. That training will now be standard.
Second, city officials can learn first-hand about regulating a technology that will re-shape mobility patterns in the next decade. The Director of Transportation Planning noted that traffic could drastically improve if AVs operated as a shared fleet with no deadheading. A few companies, like General Motors or Optimus Ride, would deploy fleets in the Metro West, renting out rides the way Uber and Lyft do today. The city or state, which would be able to monitor both the location and occupancy of AVs, could tax the companies on a scale that promotes driving with higher occupancy and parking where there is lower congestion.
The converse of that efficient scenario, though, could arise if AVs were individually owned and operated. Traffic would worsen if parking regulations did not deter owners from having their AVs circulate, with zero occupancy, to avoid paying for parking.
The third way that Newton benefits is by burnishing its reputation as a forward-thinking city in a progressive region. MIT sends the overwhelming majority of its AV engineers to the West Coast. That’s a missed opportunity for Greater Boston: we should be a hub for this fast-growing and high-paying industry.
The biggest risk is human error. Testing standards assume that having humans behind the wheel of an AV, ready to assume control, is safer than either fully human or fully autonomous driving. That logic seems spurious to me: many humans get distracted even when they are driving – how many of us could stay fully alert for many hours behind a wheel we’re not even allowed to touch?
And, even if the human is diligently paying attention, the reaction time to (1) process an anomaly, (2) physically engage with the wheel and brakes, and (3) maneuver to safety seems like it would be longer than for either a fully human or fully autonomous operator. Are we fooling ourselves into thinking that we’re in a Goldilocks zone of safety for Man + Machine, when we’re actually setting up humans to fail? This risk should be re-evaluated as testing gets underway.