Artificial Intelligence is now being utilized as the program to command machines that replaces humans in performing simple to complicated functions.
Tesla is developing a new "beast" Artificial Intelligence or AI neural computer network dubbed as "Dojo" to train electric vehicle's autopilot FSD or Full Self Driving found on Tesla Motors, expected to be fully capable by year's end.
Tesla CEO, Elon Musk (@elonmusk), announces "Dojo" on Twitter that reveals its function and uses for the electric vehicle manufacturing company. Tesla's "Dojo" is a neural network training computer that is a "beast" in nature, as stated by Musk.
Android Rookies report that "Dojo" will process vast amounts of video data integral for the Autopilot system to work accordingly. Musk also crowdsources for software engineers and developers that can join Tesla's team in procuring the program intended for the autopilot system. The job opening for the "Dojo" is located in Tesla's central engineering locations, namely: Palo Alto, Austin, and Seattle. However, according to Elon Musk, exceptional applicants will be amenable to work remotely, which will possibly be a home set-up.
Dojo Set-up
Tesla's Dojo will procure an accelerated training for the current Autopilot and FSD functionalities of the company's electric vehicles. Clean Technica notes that Dojo will be under the form of a supercomputer that is yet to be made called the exaFLOP. This enables massive loads of data processing and computations to be done in just a second. Dojo would be part of an exaFLOP-enabled computer to produce a system that will be capable of processing and functioning on various traffic situations that can drastically change in seconds. Dojo's inputs as a Neural Network for Tesla's soon-to-be 4D imaging for their autopilot and FSD would deal with a significant improvement in the electric vehicle's safety and motion. Tesla CEO Elon Musk notes that this update won't be a "tweak" or even a huge upgrade on the vehicle's self-driving system, rather a "quantum leap." The Dojo's integration is an architectural rewrite of the current system.
Tesla's Autopilot 4D
In a statement made by Musk over social media platform, Twitter, in July, Tesla's autopilot will soon see an upgrade to 4D image infrastructure from its current "2.5D," according to Teslarati.
A Twitter user and Tesla vehicle owner express his concerns about a function present in the vehicle that prevents the user from using it because of the current technology available. Tesla's Summon is a function that allows users to have their electric cars come to them, by just holding "COME TO ME" in the smartphone application.
The vehicle would then locate the smartphone that triggers the command through GPS Vectoring and proceed to go on-location. The user's concern is his driveway being ten percent more than the intended slope the vehicle is capable of commandeering. Musk's tweet provides an insight into Tesla using 4D infrastructure for its full self-driving feature on its vehicles. The use of 4D imaging could lead to improvements in the vehicle's current autopilot system to know more of its surroundings and calculate its actions accurately to suit the present situation upon drive. Elon Musk mentioned in his tweets that an updated "bleeding edge alpha built" version of the vehicle's autopilot system would be available in six to ten weeks, which is roughly more or less than two months.
ALSO READ: Apple Mac Malware Warning: 'An Unusual Infection' on Xcode Projects Compromises 380 IP Addresses
This article is owned by Tech Times
Written by Isaiah Alonzo