A couple of weeks ago, Tesla’s CEO Elon Musk teased that there is a major upgrade to Tesla’s autopilot system in the works. Given that the autopilot feature is currently facing a bit of criticism due to the various accidents that have happened as a result of using it (there is debate where some of it was user error or the system being inadequate), we guess it made sense for Tesla to want to refine the feature further.
For those wondering what the update could contain, Tesla has since officially announced the upgrade which according to the company, will now rely more on its radar system. According to Tesla, the radar system was introduced in 2014 and was designed to act more as a supplementary sensor to the primary camera and image processing system.
The reason for it being a supplementary sensor was because of the numerous issues associated with radar detection, such as how people and non-metallic objects appear to be translucent to radar, and how metallic objects such as a soda can would be misinterpreted as a large and dangerous obstacle.
However Tesla believes that they have solved this in a software update that would take advantage of a geocoded whitelist of objects, as well as through “fleet learning” where data from other cars would be reported in order to reduce false positives. At the end of the day autopilot should still not be treated as a self-driving system, something that Tesla was quick to adjust in their ads in China, but hopefully with the software upgrade, crashes that occur with autopilot will be reduced.
Filed in . Read more about Tesla.