Pesky pedestrians and bolshy bicyclists will stop driverless cars in their tracks, believes transport journalist Christian Wolmar. He calls this algorithm-enforced stasis the "Holborn problem" – when the London area fills with pedestrians exiting the busy underground station in rush hour they spill on to the road, and such an exodus would halt driverless cars, making them useless in peak periods.
Wolmar thinks driverless cars – known by wonks as "autonomous vehicles" – will be useless outside of peak times, too. He has written a short book about this most hyped of subjects, and pours cold water on those who believe AVs will ever displace human-piloted road vehicles.
"Driverless Cars: On a Road to Nowhere" was published last week and it’s the subject of a BikeBiz-connected podcast conversation recorded yesterday.
Wolmar doesn’t believe AVs will replace human-driven cars because the motor industry is still producing such cars, and the only way AVs could truly work is if they were the only vehicles on the road. If human-driven cars were on the same roads as AVs there would be lots of crashes, and there would be risks of "highway robberies" as human drivers boxed in AVs – which would have to stop – and robbed the occupants.
Similarly, AVs can’t be on the same roads as pedestrians and cyclists because they could walk and cycle in front of the safety-first vehicles.
Introducing a concept driverless car at a motor show Renault chief executive Carlos Ghosn told CBNC last year that the arrival of the technology could be delayed by cyclists who, he said, "don’t respect any rules usually."
Ghosn worried that driverless cars have a cycle-shaped hurdle to leap: "One of the biggest problems is people with bicycles. The car is confused by [cyclists] because from time-to-time they behave like pedestrians and from time-to-time they behave like cars."
(A cyclist hanging at a US intersection on a fixed-wheel bicycle in 2015 found that a Google driverless car in front of him wouldn’t budge because of his nudging to and fro in order to balance.)
Marjan Hagenzieker, professor of road safety at the Delft University of Technology in the Netherlands, told a Dutch news site in 2015 that the unpredictability of cyclists "throws a spanner in the work of technology optimists" because "robots are not good at dealing with inconsistent behaviour."
Savvy cyclists, who know they will be safe in front of a computer-controlled car, will have little incentive to cede priority. "Some will be careful with an autonomous car in traffic; others less so in the expectation that the car brakes by itself," said Hagenzieker
The Netherlands is the country best adapted to welcoming the advent of AVs, says a recent report. KPMG’s "Autonomous Vehicles Readiness Index 2018" was authored by the company’s global head of infrastructure Richard Threlfall, a former private secretary to the UK transport secretary.
Threlfall told BikeBiz: "AVs should in theory allow road lanes eventually to become narrower, because AVs should be able to navigate much closer together. That in turn should free up space which is currently assigned to [motor] vehicles, creating the possibility of much more extensive networks of cycle lanes being created alongside carriageways."
One of the conclusions of the podcast conversation with Wolmar is that AVs will need their own dedicated tracks, but for that to happen there would also have to be dedicated tracks for other road users. So, if cities were truly expecting the future to be one involving driverless cars they would be already fencing off footways and creating cycleways. How about that should be the pre-condition? Before planners and policy makers get too starry-eyed about AVs they should first create enhanced footways and super-wide cycleways.