Elon Musk Unveils Humanoid Robot and AI Chip On Tesla AI Day

Tesla CEO Elon Musk said on Thursday that the company is building a humanoid robot and is expected to introduce its prototype by next year. He said that the robot is intended to be “friendly.” Musk also revealed the specification of the Tesla bot. He revealed that the robot will be “five feet, eight inches tall, 125 pounds, and have a screen for a face.”

The company’s website reads that the Tesla bot is “capable of performing tasks that are unsafe, repetitive or boring. We’re seeking mechanical, electrical, controls, and software engineers to help us leverage our AI expertise beyond our vehicle fleet.” Musk revealed a human in the robot custom as its humanoid robot on the event of AI day. However, will Tesla Robot would be able to see the light of the day? For instance, in 2019, he said that the company is working on robot taxis that will be conquering the roads by 2020. In the same year, he said that Tesla will be introducing solar to the supercharger stations.

During the event, he projected technical details of the robot. The screen shows that the humanoid robot will have 40 electromechanical actuators, force feedback sensing, and two-axis feet for balance. He also said the code-name of the robot is “Optimus.” Other slides revealed that the robot can carry 45 pounds and lift about 150 pounds. If the company’s concept turns into reality, it can help Tesla to leverage its manufacturing units. However, Musk said that the robot is not here to our leverage our manufacturing units, but to develop computers required for robotics. “It should be able to, you know, please go to the store and get me the following groceries, that kind of thing,” he said.

Tesla Reveals Chips to Train AI Models

On the same day, Tesla unveiled its in-house develop chip to train AI networks inside its data centers. “We are assembling our first cabinets pretty soon,” said senior director of Autopilot hardware at Tesla, Ganesh Venkataramanan. Intel, Nvidia, and Graphcore are some of the companies that develop AI training chips.  The chip will help the company to develop Tesla’s self-driving system. The senior director revealed that the D1 chip uses a 7-nanometer manufacturing process. He also said that the chip will have 362 teraflops of processing power.

Musk that the chip will be in operation by next year. Currently, Tesla offers a full self-driving capability add-on to its electric cars at USD 10,000. This will help the car to automatically change the lanes and move out of the parking car without any driver assistance. Earlier, Tesla introduced the add-on at a monthly subscription of USD 199. Last year, the company had introduced a computer chip, developed by Samsung. The D1 chip production is expected to combat chip shortage and improve AI performance.

“We can do compute and data transfers simultaneously, and our custom ISA, which is the instruction set architecture, is fully optimized for machine learning workloads,” said Venkataramanan at the AI event.  The announcements came amidst the controversy surrounding its Autopilot system. Earlier this week, the National Highway Transportation and Safety Administration (NHTSA) opened an investigation against Tesla’s autopilot capabilities. It said in a public document that since 2018, it has encountered 11 crashes involving Tesla with its autopilot system.

 In a letter, Senators of Massachusetts and Connecticut have asked Federal Trade Commission (FTC) to open a probe against Tesla.  “Tesla drivers listen to these claims and believe their vehicles are equipped to drive themselves — with potentially deadly consequences,” they mentioned in the letter. Senators added that even though the name implies “Full Self-Driving,” it’s actually not. “In light of these concerns, we urge you to swiftly open an investigation into Tesla’s repeated and overstated claims about their Autopilot and Full Self-Driving features and take appropriate enforcement action to prevent further injury or death as a result of any Tesla feature.”