Uber users in Pittsburgh can now hail a self-driving vehicle to drive them around the city, in the company's push to secure a lead in autonomous driving.
The experiment is essential to the ridehailing service as it aims to grow its fleet to include autonomous vehicles.
Uber's engineers retrofitted Ford Fusion sedans with radars, light-mapping systems, cameras and sensors to give them the necessary tools for self-driving. Until the technology gets thoroughly tested, Uber asks an employee to be ready to take the wheel should the car's autonomous driving system make a mistake.
"I really believe that the most important thing that computers are going to do in the next 10 years is drive cars," says Anthony Levandowski, who is at the helm of Uber's self-driving car arm.
Uber Testing Self-Driving Cars In Pittsburgh
Uber trained some of its drivers specifically for the task of assisting the Pittsburgh autonomous cars. Passengers will be notified about the possibility of being randomly assigned to a self-driving vehicle if they hail an UberX ride. For the time being, Uber self-driving car rides are free.
While operators have to keep a light grip on the steering wheel, passengers can watch ride info from the comfort of the backseat via touch screens.
Pittsburgh is an ideal testing ground for autonomous driving as it features a wide variety of elements, from bridges and railroads to pedestrians and bicyclists' tracks, to urban driving and notoriously bad weather. The latter is one of the worst banes of self-driving cars, and testing in such conditions is mandatory for the safe release of a full fleet, as Uber aims to do.
The city is also where the company deployed its Advanced Technologies Center (ATC), a headquarters for experts in self-driving cars. Some time ago, Uber was chastised for poaching research staff from Carnegie Mellon University.
The head of Uber's ATC, Raffi Krikorian, points out that if his company can "master driving in Pittsburgh," chances are high that it can replicate the experiment in most global urban centers.
Gray Skies Over Autonomous Rivals
The past few weeks have brought the field of autonomous vehicles in the slow lane.
Recent reports suggest that Google's self-driving car project is losing traction compared to its rivals, and Apple is purportedly refocusing its efforts and axing jobs in the process.
Tesla Motors is working hard to revamp its Autopilot software system, as the first deadly crash in a self-driving car featured a Tesla Model S in May. The accident raised eyebrows and caused regulators to look into the company's self-driving technology.
All the major automakers are approaching the area with different levels of autonomy.
The biggest challenge for total driving autonomy is to feed cars with the information and understanding of roadmap scenarios that a human driver would process nearly instantly. Another problem for the future of self-driving cars is that the infrastructure is little to nonexistent, thus making the communication between vehicles and road infrastructure scarce and unreliable.
Evan Rawley, a Columbia Business School professor who specializes in ridesharing, notes that the recent hiccups for Apple, Tesla and Google are a "reality check" for near-term hopes of self-driving cars.
A few days ago, experts voiced some concerns over the safety of Uber's self-driving taxi.