I was waiting for an Uber and the driver zoomed past me at 35 miles an hour in downtown SF. (Why you wouldn’t drive slower when you’re approaching a passenger is beyond me.) When I flagged him down on the one way street, he backed up in heavy traffic to get to me.
That is the kind of driver that shouldn’t be on Uber’s platform. When I reported what happened, Uber took note and said they were going to reach out to the driver.
But the current process requires active input from passengers. Unless a drive is egregiously bad, most people wouldn’t bother.
Technology provides an easy answer to the problem: passively tracking driver behavior. If someone has a lot of quick stops, swerves a lot, brakes hard or speeds down city streets, that person shouldn’t be on the Uber system.
There’s also another huge advantage: you can track driver behavior when they don’t have a passenger but are logged on the system. The current Uber model provides a strong incentive for drivers to be reckless when they don’t have a passenger.
For the level of detail that Uber needs to bump bad drivers off the system, the sensors on modern phones are great.
There are technologies that provide even better data. A start up called Automatic sells a device that consumers can plug into the OBD-II port on 1996 or later cars. The app warns you when you are speeding and braking hard. it also provides logs of your driving and average miles per gallon.
My auto insurance is from Metromile, which uses data from the OBD-II port to charge me by the mile instead of typical pricing models. (I estimate that I’ll save 30% off my former GEICO rates.) Incidentally, Metromile won’t provide coverage while you’re providing taxi services with Uber.
One survey respondent suggested deactivating the Uber app while the vehicle is in motion, much like some embedded navigation systems. Unfortunately, the app is so dependent on people interacting while driving that that won’t happen.
Many taxi drivers are reckless, too. But we can’t do much about that.