Back in the 2018 Annual Shareholder Meeting, Tesla CEO Elon Musk noted that his vehicle, which was running a developer mode of Autopilot, was able to perform a highway on-ramp-to-off-ramp maneuver. Musk announced then that an update introducing Autopilot’s capability to enter and exit freeways would be coming sometime in the “next couple months.”
While Software Version 9’s initial release to advanced early access users proved closer to three months, the update did include a new Autopilot feature called Navigate on Autopilot, which features the capability to perform overtaking maneuvers while intelligently suggesting lane changes. Perhaps even more notable is Navigate on Autopilot’s capability to “automatically steer toward and take the correct highway interchanges and exits based on (the driver’s) destination.”
Several videos have been shared of Teslas using Navigate on Autopilot, with some even showcasing how Elon Musk’s fondly-named lane change setting, “Mad Max” mode, actually works. Just recently, owner-enthusiast Jasper Nuyens shared a video of him using Navigate on Autopilot on a highway, and while a good part of his film was more of a review of Autopilot’s settings, it was the end of his video that stood out.
Without disengaging Autopilot, the electric car smoothly exited the highway, adjusting its speed and seemingly using regenerative braking to smoothly come to a stop at the very end of the off-ramp. Autopilot only disengaged when the vehicle did not detect the driver’s hands on the steering wheel.
This recent demonstration of Autopilot’s off-ramp capability is the clearest look yet at the upcoming feature of Tesla’s driver-assist system in the Version 9 update, which is still consistently being developed and improved. Tesla is in a constant effort to push Autopilot’s capabilities forward, and so far, the company’s commitment to AI and deep neural networks appear to be paying off.
Tesla’s Autopilot system uses a suite of 8 cameras, 12 ultrasonic sensors, and a forward-facing radar to detect objects and obstructions on the road. These are paired with Tesla’s vision and neural net system which enable the company’s fleet to continuously learn and improve, based on software trained from billions of miles of real-world data gathered by its cars. With every iteration of Autopilot, Tesla’s neural network gets more advanced. Thus, in the future, it would not be surprising if Autopilot-driven vehicles would stop at the end of a highway off-ramp because it recognized road signs.
In its continuous effort to improve its driver-assist system while preparing for the rollout of the first features of its Full Self-Driving suite, Tesla is involved in the creation and development of its own custom hardware. Details of these chips were teased during the company’s Q2 2018 earnings call when Elon Musk discussed some aspects of Hardware 3. Musk proved optimistic about Tesla’s custom silicon, stating that it would be the world’s most advanced computer that’s designed specifically for self-driving.
“It’s an incredible job by Pete and his team to create this, the world’s most advanced computer designed specifically for autonomous operation. And as a rough sort, whereas the current NVIDIA’s hardware can do 200 frames a second, this is able to do over 2,000 frames a second and with full redundancy and fail-over. So, it’s an amazing design, and we’re going to be looking to increase the size of our chip team and our investment in that as quickly as possible,” Musk said.
Watch Tesla’s Autopilot perform a highway off-ramp maneuver in the video below.