Just as the first fatal crash in a Tesla with the Autopilot activated came to light last week, we revealed that the automaker is planning to release its software version 8.0. Tesla refers to the update as the most important to its touchscreen since the launch of the Model S in 2012, but it will also feature significant improvements to the Autopilot.
While Tesla has obviously been working on the update before the highly publicized crash, which happened in May, the timing is still interesting as Tesla’s semi-autonomous system is now subject to more scrutiny than ever before.
The new 8.0 update, which is currently being tested with a select few Tesla owners and should be pushed to the fleet in the next few weeks, will allow for improvements in the general Autopilot experience in traffic, but more significantly, it will introduce automatic off-ramp in exits on the highway.
When Elon Musk first introduced the Autopilot, he made it clear that hardware limitations will not allow for a fully self-driving system, it will have to wait for the Autopilot 2.0 (more on that soon on Electrek), but he said that the system will enable self-steering on highways and self-driving at low-speed in parking situations, like with the Summon feature.
As evidenced by the tragic reminder that came in the form of the fatal accident revealed last week, the system needs to be monitored at all time and the driver needs to be ready to take control, but yet the Autopilot was presented as enabling driving on the highway “ramp to ramp” without having to touch the steering wheel (to correct the trajectory – Tesla still recommends keeping your hands on the wheel).
The v8.0 update will truly enable this vision by allowing off-ramp exit through activating the turn signal – much like the automatic lane change under the current version of Autosteer.
A common problem with the early version of the Autopilot was that it had a tendency to try to take exits on the highway when it wasn’t supposed to. After just a few weeks, Model S owners fed the Autopilot’s collective network intelligence with enough information that it was able to largely eliminate the issue.
The Autopilot learns from all the vehicles equipped with the hardware in Tesla’s fleet (~80,000 vehicles) by building high precision maps, which it refines with every passing of a vehicle, and then downloads the map sections aligned with the vehicle’s GPS to help the vehicle’s own Autopilot system navigating the location in real-time with cross-checks from the vehicle’s sensors, primarily its front-facing camera and radar.
The more the fleet travels, the better the Autopilot will get.
With the 8.0 release and the addition of automatic off-ramp exit, the Autopilot will be getting a lot closer to the vision Elon Musk described for the first generation of the system. While there will certainly still be room for refining the system, it’s difficult to imagine many new features being implemented at this point.
This is where Autopilot 2.0 comes in. A new sensor suite on which Tesla can release more advanced autonomous and semi-autonomous features built on the architecture developed through the current Autopilot program.
At this point, Tesla collected about a billion miles of data with its vehicles equipped with Autopilot hardware and around 130 million directly using Autopilot (Tesla said 780M miles of data, 100M miles driven last month). The data will become increasingly useful if it can be cross-checked in real-time with more sensors, which could in turn potentially increase the level of safety by a significant amount.
For more details on the 8.0 update, you can read our full exclusive report from last week: Exclusive on Tesla 8.0 update: new Autopilot features, biggest UI refresh since launch and much more
Featured Image: A member of the media test drives a Tesla Motors Inc. Model S car equipped with Autopilot in Palo Alto, California, U.S., on Wednesday, Oct. 14, 2015. Tesla Motors Inc. Photographer: David Paul Morris/Bloomberg
Subscribe to Electrek on YouTube for exclusive videos and subscribe to the podcast.