Joining Forces: How Driverless Cars are Bringing Ride-Hailing Services and Auto Manufacturers Together

by Pengkai Xia

When I heard that Audi was testing an autonomous RS7 on the iconic Hockenheim motor racing circuit [1], I found myself somewhat disappointed. As a car enthusiast, driving a RS7 has been on my sport cars wish list for years; I had dreams about getting the chance to drive it on the track and feel the force from that twin turbo V8 in real life. That kind of feeling and experience is what car lovers like myself live for. With the shift to autonomous, I would become just another passenger.

Putting my disappointment aside, I soon realized Audi was doing what was necessary to keep up with the constantly evolving auto industry. They were focusing on calibrating their autonomous technology to teach computers to drive fast safely — a step forward to bring sophistication to self-driving cars. This move comes right out of the ride-hailing market — they are leading the charge in testing autonomous concepts and car companies have decided to join the movement and solve the problem together.

The Challenge

One of the biggest challenges for autonomous vehicles is precision. Equipped with cameras and sensors, computers are capable of driving the car in a controlled environment where the road is clean and the weather is pleasant. But in reality, we all know that road conditions are never reliable and sunny days are never a guarantee. With this, computers must learn how to make the right decisions in unpredictable situations and prepare for the unexpected.

http://www.wired.com/wp-content/uploads/2014/10/RS7c140029-1024×768.jpg

But driving is a learned process. Teenagers spend hours driving with parents, then sitting through drivers-education and then testing before they are ready to tackle the roads alone. Most professional racecar drivers take years racing in go-karts before their skill level is ready to participate in tracks like Formula One. Therefore, it is just as crucial for the computers to get just as much experience with as many road or weather conditions as possible before they are able to generate a system that can make decisions on its’ own.

Achieving this and getting consumers on board with these autonomous cars has not been an easy road (pun intended :p). Manufacturers have covered millions of miles around the world on their autonomous test models, yet improvements are limited. While these tests have helped computers recognition in various conditions and consolidated high-resolution 3D map development, computers still have to learn how to make decisions in unpredictable situations from humans. In order to achieve this, manufacturers need to conduct their tests in a platform that carries large volume of people who use cars frequently…

Enter, ride-hailing.

The Solution

Currently, raid-hailing services are utilizing a “driver with car” recruiting model, but this comes up short on actual drivers that can operate the service. Lyft estimated that roughly 150,000 [3] drivers who wanted to sign up for Lyft did not have a qualified car to operate the service. Similarly, both Uber and Lyft have been facing a variety of lawsuits and negative press around the actual drivers in the car. While the service is popular, it has a high-level of risks and costs.

This has turned all ride-hailing services towards autonomous cars. Lyft, Uber and others are excited by the efficiencies that utilizing autonomous cars could create — both by having more control of the driving process and by lowering labor expenses with less actual drivers on the roads.

https://fortunedotcom.files.wordpress.com/2016/03/express-drive-lyft.png?w=921

But before your next Uber drives itself to pick you up on a night out, there’s still a lot left to figure out. To help achieve this, GM created a partnership with Lyft that would help both sides get closer to solving the puzzle of autonomous cars; GM agreed to rent its Chevy Equinox to Lyft drivers [2], providing Chevy with access to driver data and Lyft drivers with safe and reliable cars.

There are also speculations that GM’s motivations in this partnership stretch further, perhaps to being a stepping stone for GM’s own car-sharing service Maven, which will be aiming to take car connectivity to the next level.

Sounds like a win-win for everyone and I’m excited to see what comes out of it. But as the token car guy in the office, you can bet I’ll still be driving my own car for years to come.

http://blog.caranddriver.com/wp-content/uploads/2016/01/GM-Maven-Car-Sharing-interior-626×361.jpeg