Self-Driving Cars & The City
Last week, I attended the first edition of Sidewalk Talk: Beer & Brainstorm at Hudson Yards, hosted by Sidewalk Labs in their brand new office. The topic du-jour was « the role of cities regarding autonomous vehicles (AV).»
My Take aways
- 94% of road kill casualties are due to human error. This is a public health problem and AVs can really save lives.
- Legislation needs to change to permit AVs. For instance, in New York State, a 1971 measure requires drivers « to keep at least one hand on the steering wheel while the vehicle is in motion. » (source)
- How good is good enough for AVs? Humans are far from perfect. Why should we ask AVs to be perfect too? Isn't it enough if they are better than human drivers?
- AVs are a great solution to solve the parking space problem.
- We should think more creatively about AVs in terms or size and shape, e.g. small pizza-delivery car with built-in oven for baking on-the-go.
My Thoughts
Listening to conversations about autonomous vehicles, I see a few distinct concerns:
- a legal concern, mainly about responsibility. In case of a crash, who should be blamed and who should pay?
- a safety concern, about how ready the technology is and how easily it could be hacked by evil-doers
- an ethical concern, about algorithmic decisions made by AVs (read about the Trolley Problem if you have never heard of it)
- an equity concern, about people losing their jobs and who will benefit first from the technology.
All of these concerns are valid. But we have to remember that before it is deployed, AV technology needs to be tested and bootstrapped. For this transition period, I feel that (1) and (4) should not take precedence.
My Radical (and probably naive) Solution to Bootstrap AVs :-)
Thinking about the problem of introducing AVs into urban environments reminded of the following anecdote: how do you boil a frog?
You have at least two options to boil a frog.
- You put water in a pot; you bring the water to boiling temperature; you drop the frog in the pot.
- You put water in a pot; you put the frog in the pot; you slowly bring the water to boiling temperature.
The problem with option 1 is that the frog might guess your evil intentions and attempt a last second salvation jump when it hits the hot water.
Since humans make mistakes, the road/street infrastructures have to be designed to tolerate them. Only 5% of pedestrians die when hit by a car going at 20 mph compared to 40% (at 30 mph) and 80% (at 40 mph). Speed reduction is therefore a key tenet of initiatives like Vision Zero, aimed at reducing pedestrian casualties.
So, here is a proposal for autonomous vehicles that should address the safety concern
- pick a speed limit at which AVs cannot do any harm
- deploy AVs, learn and make things better
- keep increasing speed as needed
In Lyon, France, the recently introduced (October 2016) autonomous bus NAVYA ARMA cruises at 25 km/h to guarantee the safety of passengers and pedestrians (source: NAVYA ARMA press kit). Of course it is too early to tell if and how it also successfully addresses the other concerns.
Conclusions
Autonomous vehicles are a very promising technology that should save lives, reduce the parking space problem and probably make cities more equitable by providing affordable and shared means of transportation.
Therefore, if it is not already, autonomous vehicles (AVs) MUST BE on the radar screen of city planners.
A few extra interesting resources mentioned during the event:
