Self-driving vehicles start with Red Whittaker.
The Carnegie Mellon University robotics professor was a fixture at the DARPA challenges held more than a decade ago, which helped spark the modern-day push to build autonomous-driving systems.
In 2007, he outfitted a Chevy Tahoe known as Boss with sensors, software and computing power in an architecture that has been a blueprint for dozens of companies building such systems today. Boss won the Urban Challenge that year, bringing Carnegie Mellon a $2 million prize.
For Whittaker, it's one of many highlights in a remarkable robotics career. He joined Deputy Mobility Editor Pete Bigelow and Editor Leslie J. Allen on the Shift podcast to reflect on the DARPA challenges, the progress of self-driving technology and his latest moonshot. Here are edited excerpts.
Q: Why did boss fare so well?
A: The Boss was a ... superb autonomous racer. It had all the sensing you'd come to expect from such a car today, and server-class computing of the time. It had tremendous capability for a plan A and plan B, which you needed because racing is so unpredictable. It had so much more than what was needed to bring in the win.
Why Did you choose the Tahoe? because it was the best vehicle for the desert?
(Laughs.) It was chosen because General Motors paid the bill. The challenges had nothing to do with high speed. If all were known, on the morning of the Urban Challenge, DARPA approached me with a clear request to tone it down, cut it back, no sense to overdo it.