In a major milestone for robocars, Waymo has announced they will deploy in Phoenix with no human safety drivers behind the wheel. Until now, almost all robocars out there have only gone out on public streets with a trained human driver behind the wheel, ready to take over at any sign of trouble. Waymo and a few others have done short demonstrations with no safety driver, but now an actual pilot, providing service to beta-testing members of the public, will operate without human supervision. This is a big deal, and indicates Waymo's internal testing is showing a very strong safety record. The last time they published numbers, they had gone 83,000 miles between "required interventions."
Although other companies that are working on autonomous driving might get more attention, Waymo is still hard at work on the technology. The Alphabet subsidiary just provided an update on its Waymo Driver AI as well as more details about its self-driving tests. An array of LiDAR, radar and cameras can track what's going on all around the vehicle in a variety of weather conditions, Waymo says. The system generates a 3D view of the vehicle's surroundings that humans would be able to understand. Along with other cars, the system can render pedestrians in addition to cyclists who narrowly pass by the vehicle.
The other very common type of disengagement is a software disengagement. Here, the software decides to disengage because it detects something is going wrong. These are quite often not safety incidents. Modern software is loaded with diagnostic tests, always checking if things are going as expected. When one fails, most software just logs a warning, or "throws an exception" to code that handles the problem. Most of the time, that code does indeed handle the problem, and there is no safety incident. But during testing, you want to disengage to be on the safe side. Once again, the team examines the warning/exception to find out the cause and tries to fix it and figure out how serious it would have been.
A link has been posted to your Facebook feed. Waymo self-driving cars are seen Nov. 28, 2018, in Chandler. Emergency crews directed the afternoon traffic around the wrecked cars and fire engines at McQueen and Pecos roads in the Phoenix suburb in mid-October. The Chrysler Pacifica minivan -- equipped with former Google car company Waymo's self-driving vehicle technology -- approached the scene tepidly, while dozens of other vehicles merged into the turn lanes far sooner. A human driver in this situation might try to make eye contact with the drivers already in the crowded turn lane, or even wave, to try to cut in.
Waymo is reportedly rolling out additional safety measures for its self-driving vehicle fleets, reintroducing safety drivers and installing cameras to monitor driver fatigue. The Information reports that these changes were put into place due to safety concerns, and they come after a handful of recent traffic incidents. Within the last few weeks, Waymo has put safety drivers back behind the wheels of its more advanced vehicles, which have been operating without such drivers for some time, and across its broader fleet, the company has added co-drivers to its daytime shifts as well as its night time shifts. The co-drivers are part of Waymo's effort to keep its safety drivers alert, and The Information reports that the company has also been installing cameras aimed at drivers' faces for the purpose of monitoring when they might be nodding off. In June, a safety driver appeared to fall asleep while behind the wheel of one of Waymo's Pacificas, causing an accident after he inadvertently turned off the driving software.