Robo-cars? Portland’s got a plan for that.
But does it truly serve the public?
As first reported by BikePortland, Portland is trying to be “open for business” for autonomous vehicles. But how smart is Portland’s smart car plan? Let’s consider the four actions Portland’s Mayor is directing the Portland Bureau of Transportation (PBOT) to take by mid-June :
1. Propose for City Council and public consideration Interim Transportation System Plan (TSP) policies that ensure connected and autonomous vehicles will serve Portland’s safety, equity, climate change, and economic goals;
2. Publish a Request for Information (RFI) that invites AV testing specific to advancing safety, equity, climate, and economic goals;
3. Adopt an Interim Administrative Rule that provides a clear path to permit innovators to apply to test, pilot or deploy AVs in Portland; and
4. Develop public engagement, reporting, and evaluation plans that ensure Portland residents, workers, and businesses have opportunities to shape the “rules of the road” for AVs in Portland.
The city, whose reputation for planning is second only to its “green” cred, is smart to call out “safety, equity, [and] climate change” as leading concerns, so points 1 and 2 are good.
Points 3 and 4, however, leave something to be desired.
Autonomous vehicle companies need complex testing environments to develop and train their creations, and by using publicly-funded streets — full of other people’s cars, bicycles, children, etc — they stand to save a great deal of money by not funding the construction of such “real life laboratories” on their own.
This is a huge gift to them, and it seems the leadership at City of Portland realizes the public can and should get some public benefit in return–that this is a chance to ensure vehicles of the future serve our goals for improvement on “safety, equity, climate change, and [the economy]”.
We should be concerned about timing on points (3) and (4) of that directive. Point (3) calls for an interim rule “that provides a clear path to permit innovators to apply to test, pilot or deploy AVs in Portland” while point (4) directs PBOT to “Develop … plans” so Portlanders can help ‘shape the “rules of the road” for AVs in Portland.”’
If PBOT takes these actions simultaneously “within 60 days”, then at the end of the 60 days the permit applications would be available to companies but no actual public input would have been received on how the testing itself will happen or what those permit applications contain.
The directive appears to indicate that the city is interested in public input for the eventual “rules of the road” that the eventually-tested autonomous vehicles will operate under, but does not explicitly indicate the city will seek public input on the “rules of the road” for the testing process the vehicle companies will operate in.
Portland’s city leadership should model their testing process along the lines of California, which first published draft testing regulations and then asked for public feedback on those regulations before implementing them. (Those comments were recently published online.)
California DMV Asks Public To Weigh In On Self-Driving Car Rules
SACRAMENTO (CBS / AP) - California regulators asked members of the public Wednesday what they thought about proposed…
There are a number of considerations related to testing autonomous vehicles that should be evaluated before granting a permit for testing: the types of sensors the vehicles have, the data they collect, the datasets the vehicles need to operate like navigational maps or LIDAR environment maps, the robustness of the code that operates the vehicle, etc.
And there are a number of ways the public could benefit from the testing process itself, if it is designed well. For example, would you want four private companies driving multiple fleets of LIDAR mapping cars around Portland to build up separate private datasets? Or would you rather have the LIDAR mapping done once to create a public dataset. A LIDAR map is a highly accurate 3D map of the built environment, produced by scanning that environment with lasers.
Such a public dataset could serve a multitude of government and civic purposes: it could be used:
- by the city to quickly assess the condition of all of the city’s street trees
- by the autonomous vehicle companies to develop their vehicles
- by academics examining the safety of autonomous vehicles in simulated environments
- or by open source developers working on open source autonomous vehicle driving platforms which might power a future TriMet self driving shuttle, for just one example.
Much of the data generated or required for the running of these ‘smart’ vehicles (LIDAR, maps, images, vibration sensors that detect curbs of sidewalks while parking or potholes while driving, etc) is extremely valuable and costly to acquire otherwise.
Autonomous Cars Will Require a Totally New Kind of Map
As the vehicle navigated the labyrinthine streets of London and headed for the countryside of Surrey with uncommon…
The City should think of data as the public infrastructure of its digital future — it’s data like this that can make Portland a truly “smart city” which serves everyone. This new digital urban frontier should serve not just Millennials hailing self-driving Ubers programmed by brogrammers in Silicon Valley, but also the 8 year old in a wheelchair in East Portland trying to get to a park who wants to ask her smartphone for a route known to have good sidewalks and curb ramps (thanks, Smart City digital infrastructure).
At the very least, in the event of a crash, the data collected by the car should be turned over to the City. Uber would rather not.
Obviously, the data these VC funded self driving car companies use and generate is valuable and no one wants to give anything away for free. But the same could be said of our streets.
Thanks for reading! — @techieshark.