Tesla is preparing to launch a new Autopilot hardware 4.0 upgrade, according to Electrek. It is expected that the automaker will announce a new Autopilot/Self-Driving hardware suite today. The hardware is referred to as Hardware 4.0.
(Photo : by Justin Sullivan/Getty Images)
FREMONT, CALIFORNIA - APRIL 20: In an aerial view, Tesla cars sit parked in a lot at the Tesla factory on April 20, 2022 in Fremont, California. Tesla reported first quarter earnings that far exceeded analyst expectations with revenue of $18.76 billion compared to expectations of $17.80 billion.
Major Changes are Coming
Some indications point to major changes that are coming. For one, the removed radar sensors will see a return as the EV maker plans to add one this month. In addition, the Tesla Model 3 prototype will potentially feature a new Autopilot/Full Self-Driving hardware suite.
There is also a report from China that showed Tesla applied with regulators for several modifications to its vehicles, such as new Autopilot cameras.
Now, the company has done the same in Europe, where they filed changes to their vehicles with regulators in the Netherlands.
The Tesla forum TFF has also revealed a new application for Model S and Model X, such as Generation 4 complex vehicle control system (Autopilot), new car computer and GNSS antenna, Gen3 inverters, and more.
This confirms that the company is set to release the new Autopilot hardware 4.0 and include a new computer.
Recently, Elon Must commented on Hardware 4.0 and confirmed that his company isn't planning to offer retrofits for older models.
Also read: Tesla Autopilot Team to Work For Optimus Humanoid Robot Project, Says Elon Musk
Concerns Over Greenhouse Gas Emissions
Tesla is getting ready to launch new Hardware 4.0 computer technology for its next version of the Full Self-Driving Beta. However, Massachusetts Institute of Technology researchers have warned that cars with autonomous driving systems could produce the same amount of greenhouse gas emissions as the world's data centers.
Suppose by 2050, 95% of the electric cars on the roads have some form of autonomous driving with the necessary hardware and sensors. In that case, the computational efficiency of these cars will have to be improved on an almost yearly basis in order to keep the emissions below the current levels of data centers.
A self-driving electric car with ten cameras and a neural network can make 21 million decisions in a single hour of driving. If two-thirds of the world's current car fleet is replaced with these types of cars, it will result in a staggering 21.6 quadrillion decisions in just an hour. This is many times more than the amount of data Facebook is able to process in a single day, even with its data centers.
Research indicates that for autonomous driving systems to have an acceptable level of emissions throughout their life cycle, their power draw should not exceed 1.2 kW.
The authors of the MIT study advocate for making autonomous driving technology more efficient in terms of energy while still ensuring the utmost safety of the vehicles. To do this, Tesla's upcoming Hardware 4.0 should be made increasingly specialized, and the algorithms used for self-driving should be improved for improved efficiency.
Related article: Tesla Builds a New Power Plant in Japan Using Powerwalls Home Battery Pack
