Tesla Vision

Tesla today announced the official transition to “Tesla Vision” without radar on Model 3 and Model Y.

In the process, the automaker warns of some limitations on Autopilot features at first.

After several changes in the Autopilot team’s leadership over the past few months, Tesla ended up hiring one of the leading AI and computer vision researchers this week.

At the same time however, we learn that several of the team’s top computer vision experts have left the company over the past few weeks.
Expand
Expanding
Close

There have been some significant comings and goings in Tesla’s Autopilot leadership over the past few months. Most significantly, Tesla hired the creator of the Swift programming language, Chris Lattner, from Apple to lead the Autopilot software team. It enabled Jinnah Hosein, SpaceX’s Vice President of Software and who had been filling the position, to get back to his regular job. Sterling Anderson, Tesla’s Autopilot program director, also left the company to start his own self-driving startup and he was subsequently sued by Tesla over his hiring of a few colleagues.

Now we learn of Tesla’s Vice President of Autopilot Vision, David Nistér, leaving the Autopilot leadership team for Nvidia.
Expand
Expanding
Close

Intel’s $15 billion acquisition of Mobileye is all the talk in the auto industry today. The biggest question is how do they value the company at 30 times its projected revenue for the year? It’s a historic acquisition in term of size and valuation. For comparison, Tesla is valued at $40B, about 4 times its projected revenue for the year.

The answer is that Intel sees the acquisition as bringing in-house the “entire package” of autonomous driving. CEO Brian Krzanich wrote in an email to employees today: “this acquisition essentially merges the intelligent eyes of the autonomous car with the intelligent brain that actually drives the car.” The eye being Mobileye and the brain being Intel.

The fact that they are willing to pay $15 billion to accomplish that is quite interesting and somewhat validates Tesla’s own new approach which aims to do same thing since discontinuing Mobileye’s system in the Autopilot. In fact, Tesla considers its in-house efforts an improvement over the Mobileye platform. 
Expand
Expanding
Close

Tesla’s software timeline to fully autonomous driving on its new Autopilot hardware can be somewhat complicated. There’s ‘Enhanced Autopilot’, which in itself offers several different features, and there’s ‘Full Self-Driving Capability’, which despite its name will not enable self-driving for a while, but could still be useful to Tesla drivers very soon.

CEO Elon Musk clarified the timeline last night.
Expand
Expanding
Close

A picture of the instrument cluster of a Tesla Model S has been circulating on Tesla forums and Facebook groups following the company’s latest Autopilot update which brought the ‘Enhanced Autopilot’ feature to new cars with the second generation hardware.

After checking with Tesla, we can confirm that a vestigial version of the feature could have made its way in a build released to the first 1,000 cars with the new hardware this weekend, but it wasn’t intended to be in the latest customer build since the system still doesn’t act on the stop signs.
Expand
Expanding
Close

Tesla didn’t want to start the new year on a bad note by missing a deadline with its Autopilot update for new cars to work is its ‘Tesla Vision’ image processing system and while it didn’t bring the system to parity with the last generation Autopilot, the company sort of kept its ‘December 2016’ goal for the release of ‘Enhanced Autopilot’, but it’s only what the automaker is calling the “first phase” of the new features.
Expand
Expanding
Close

As new owners of Tesla Model S and X equipped with the new Autopilot 2.0 hardware are growing impatient for not having the same convenience and safety features as owners of vehicles equipped with Tesla’s first generation Autopilot, CEO Elon Musk took to Twitter to reassure everyone that the company is making progress.
Expand
Expanding
Close

Tesla has been rumored to be working on its own SoC (System on Chip) optimized for self-driving cars since we reported that the company quietly hired legendary chip architect Jim Keller from AMD as new “Vice-President of Autopilot Hardware Engineering” earlier this year. While Keller’s hardware engineer experience could be useful for other projects at Tesla, the fact that the automaker poached a team of chip architects and executives from AMD following Keller’s hire fueled the rumor.

In what could now possibly be a confirmation of the rumor, a report from South Korea suggests that Samsung Electronics signed a contract with Tesla to build an ASIC (Application-Specific Integrated Circuit) system – meaning to build its System on Chip with Samsung semiconductors.
Expand
Expanding
Close

As we reported last month, Tesla’s plans to bring self-driving cars to market without using LiDAR sensors are raising a few eyebrows. That’s because we have been told for years now by most people in the industry that it wasn’t achievable, but the company moved forward with its plans in a big way last month when it started producing all cars with a self-driving-ready hardware suite without LiDAR.

The Drive is now confusing quite a few people about Tesla’s strategy by claiming that the automaker is actually planning to use LiDAR in a new article titled: “Despite Public Stance, Musk Secretly Plans to Use LiDAR in Future Tesla Models: Ignore Musk’s tweets and what he’s said in earnings calls, a Tesla equipped with LiDAR is in the works.” (Update: the Drive has since changed its headline to nuance the claim)

But it’s actually not the case. Here’s why:
Expand
Expanding
Close

The announcement that Tesla is now equipping every car coming off its assembly line with what the automaker believes to be the necessary hardware to enable full self-driving capability has been met with skepticism among industry watchers and left the market mostly unimpressed.

Tesla’s stock price fell by 2% after the announcement, which indicates that the market either has doubts about Tesla achieving level 5 full autonomy with the new hardware or it doesn’t understand the implications of having full autonomous capabilities. The latter is unlikely considering the value of self-driving technology for automakers has been mostly understood for the past few years now.

The former is more likely the case here since we have been told that lidar sensors are required for full autonomy and redundancy ever since self-driving vehicle development has become mainstream in the industry, and Tesla is almost famously not using the laser-based sensor.
Expand
Expanding
Close

We reported in exclusivity earlier this month that Tesla was going with Nvidia hardware to power its imagine processing platform called ‘Tesla Vision’, but we couldn’t pinpoint which product Tesla was going to use exactly. While we discussed the possibility of using Nvidia’s new Drive PX2 AI computing platform for self-driving cars, we also noted that the product is fairly new and expensive to be included in all new Tesla cars coming off the line.

We thought the theory was confirmed when Tesla CEO Elon Musk said this week that the automaker ended up choosing the ‘Nvidia Titan GPU’, an off-the-shelves but powerful GPU card, to power ‘Tesla Vision’, but as it turns out, Musk misspoke and Nvidia confirmed that the “in-vehicle supercomputer is powered by the NVIDIA DRIVE PX 2 AI computing platform.”
Expand
Expanding
Close

While Tesla’s announcement yesterday has a ton of incredibly interesting implications for the near future of the company and whole industries really, and we will get into those today or by the end of the week, let’s start by looking into the “product update” itself which is the addition of new hardware in all of the new Teslas rolling off the line in Fremont as of earlier this week.
Expand
Expanding
Close

Tesla was able to stop any information from leaking ahead of the product unveiling today, which makes today’s event a rare exception when it comes to Tesla unveilings and it doesn’t help to suppress any kind of hype or speculation.
Expand
Expanding
Close

It’s arguable, of course, but third-party tests have shown Tesla Autopilot outperforming other semi-autonomous or advanced driver assistance systems (ADAS) from Mercedes, Hyundai and Cadillac by a wide margin.  That, and the fact that Tesla is gaining more real world data in its vehicles than anyone else would seem to indicate Tesla is a leader in the field, if not the leader.

Now we learn that Tesla could be about to significantly increase its lead with ‘Tesla Vision’. Electrek has learned more details about the new program, which is an end-to-end computer vision system built with NVIDIA’s CUDA, a parallel computing platform by the graphics processing unit (GPU) maker.
Expand
Expanding
Close

We reported earlier today on comments made by Mobileye CTO Amnon Shashua about what is starting to look like an ugly breakup between the Israel-based maker of driver assistance systems and Tesla. He claimed that Tesla was “pushing the envelope in terms of safety” and that the company wasn’t comfortable with it.

But now Tesla is painting an entirely different picture of the reasons behind the end of the supply relationship. In doing so, Tesla confirmed its in-house ‘Tesla Vision’ product for computer vision and it depicted shady alleged business practices at Mobileye.
Expand
Expanding
Close