Google has recently released new data and findings on their self-driving car. As Google says themselves in their June 2015 report, “We’ve made a lot of progress with our self-driving technology over the past six years, and we’re still learning. Every day we head out onto public streets so we can keep challenging and refining our software.” Google also announced that moving forward, they will be releasing monthly reports.
Much of the released information is positive. For example, with the help of big data analytics tools, the self-driving cars are able to tell the difference between regular vehicles and emergency vehicles and will stop when an emergency vehicle is detected nearby. The report says, “Our car was stopped at a stoplight. Our light had turned green, but we detected an ambulance approaching from the right, so we remained stopped as it passed through the intersection.” In fact, the self-driving cars are very perceptive of all movement around them; in one case, while driving at night, two bicyclists suddenly changed paths, and one of them entered the Google car’s lane. Google reports, “Our car was able to predict that cyclist’s path of travel, so we stopped and yielded. This happened at night, when it would have been very difficult for a human driver to see what was unfolding.”
However, in this report, Google also provides detailed information for the very first time on the accidents that their self-driving cars have been involved in between 2010 and 2015. They report, “In the six years of our project, we’ve been involved in twelve minor accidents during more than 1.8 million miles of autonomous and manual driving combined. Not once was the self-driving car the cause of the accident.” Google also explains that they are disclosing data on all twelve prior accidents in this particular report, but any future accidents will be reported in the report for that month.
After analyzing Google’s reported data, it is important to note that the company is correct: each of these accidents were caused either by another vehicle or by human error while the Google car was under “manual mode” rather than “autonomous mode”. In autonomous mode, the human driver of the car isn’t touching the car’s controls at all; the software is doing all of the driving. Since each of the accidents were caused by human error, this new information can serve as an effective argument in favor of adopting these self-driving cars. But why did Google take so long to reveal this data?
The primary reason Google took so long to release this information is that, according to NBC News, the California Department of Motor Vehicles has only recently allowed for these reports to be made public. The Associated Press had “argued to the Department of Motor Vehicles that the agency was improperly withholding the information.” As the Associated Press’ spokesperson Paul Colford said, “Unlocking these records and sharing them with the public are in keeping with AP’s longstanding efforts nationwide to bring about greater transparency in government agencies.” The California DMV then “determined they could release the reports as long as personal information such as the drivers’ names is blacked out.” Google was then able to release the referenced report detailing the accidents and will continue to do so.
So what does this mean for the future of self-driving technology? Google finally releasing the details of the self-driving cars and their accidents, and the fact that they’ll be disclosing these reports on a monthly basis from now on, is actually a good thing for these types of cars. It’s reassuring that Google is going to provide all the data on these cars now that the DMV will allow it. Such full disclosure on something as new, unfamiliar, and dangerous-sounding as driverless cars is crucial in convincing the public that the invention is a good idea.
Even though it initially sounds dangerous that the self-driving cars have been in twelve accidents, all the details in Google’s report show that the cars were never at fault. It really goes to show that it’s really potentially more reliable to have a computer-driven car that can sense everything around it without getting distracted as human drivers do — and this technology, much like cloud computing, will only improve over time.
~ Rick Delgado