Tuesday, October 16, 2018

Alternative Take on AV Legislation and Regulation

Sen. John Thune and legendary consumer and safety advocate Ralph Nader are duking it out in the South Dakota press over an automated vehicle (AV) bill pending in the US Senate. I am not sure that either gentleman has expressed himself well. Thune does not provide a reason for the lack of AV safety standards or performance measures in the legislation - or mandating that the US Department of Transportation develop - and only cites the large number of fatalities on America's roads each year, while Nader gets way into the weeds on every current standard without addressing that AVs might not need the same type of federal vehicle safety standards - FMVSS - as conventional vehicles.

I'd like to offer a reasonable rebuttal to both. Meanwhile, reports change every few days about the likelihood of an AV bill passing during this session of Congress.

Scope is a big issue

Neither Thune nor Nader mention that there is already partially automated vehicle technology on the road. This is a significant issue that Congress has completely failed to address and there are no federal standards for safeguards to ensure that drivers remain attentive when a partially automated vehicle is in AV mode because while the AV mode is on, the driver might, at any moment, be called upon or find it necessary to take over. Research has indicated that we humans quickly become complacent and stop paying sufficient attention when a machine generally does its job.

Slow is fine

First, to counter Sen. Thune, there is no reason to rush because we have at the very least a few years before AVs are widely available for transit or personal use. In the mean time, we have plenty of legislative approaches at the state and international levels that we can sit back and learn from. The benefit (and, yes, the disadvantage) of our constitutional framework for passing legislation is the high level of consensus needed. What this generally means is that when federal legislation passes - yes, generally, and not in all cases - there is compromise and consensus.

US technology is in the lead without a federal statute; there will be a market for AVs whether we sooner pass a federal statute or later, so why not allow states and other countries to be petrie dishes for regulation, insurance coverage, transit applications, etc. before Congress steps in?

Issues that the federal government should place more resources into studying and addressing are cybersecurity and privacy, emergency preparedness and management, a broadened scope of accessibility for people with disabilities, transportation equity, labor and business model implications (for drivers, car dealers, equipment supply chains, etc.), roadway safety for walkers, bikers, and other modes, and geographic coverage. And I did not want to mention whether every single person will be expected to carry a smartphone at all times.

At the very least, Congress should pay attention to the Transportation Research Board's Forum on Preparing for Automated Vehicles and Shared Mobility. This advisory group issued a circular in Sept. 2018 about the many concerns that we should address at the national level as we contemplate a rollout of AV technology on our roads. I'll write about that circular in a future post because I have not read the whole circular yet.

Yes, the USDOT's AV 3.0 is out on the street and USDOT is accepting comments. I am still making my way through the 80-page document. And ditto: I'll write about that document in a future post.

Friday, October 5, 2018

Note to Ford: Great Idea for AV Esperanto

Far be it from me to compliment an auto company, but here goes: Ford is doing the world a major favor by spearheading a serious effort to establish a common language - or rather a set of symbols, colors, or images - for autonomous vehicles (AVs) to communicate TO - not with - pedestrians, bikers, and other road users (those not in vehicles). There are several reports out, but I like the detailed article posted on the Automotive News site.

I thank Ford Motor Company for its practical approach. Ford has been working on this for years and it is currently testing its pedestrian-message signals in Pittsburgh. Is Ford patting itself on the back? Fine, whatever. 

All about public acceptance

Ford understands the value of public trust. This desire for trust is a major reason why AV pilot programs - whether shuttles or car based - have low speeds and human monitors. The recent report, M City Driverless Shuttle - A Case Study, states: "Driverless shuttles have a future only if they are trusted and used by riders, and trusted and accepted by other road users." (p. 6) In fact, M City basically acknowledges that it was not technological advancement that prompted the pilot, but rather - apart from being cool and somewhat in the forefront - "understanding passenger and road user behavior while ensuring a safe deployment."

What about pedestrian-to-AV communications? 

My question to AV manufacturers and tech engineers is what to do with the humans who will continue to be in the mix. To use a crass example, when I yell out "you jerk!" to a driver, with body language commensurate to such a message, the driver comprehends what I am saying without having to actually hear me. He or she might not like the message, but it is understood. What if I find myself in an emergency and I wave for an AV to stop? Will the AV - possibly without a human passenger inside - comprehend my loud shouting and gestures?

Are companies also working with the recognition that communication must be more than one way?