Cleantech & EV'sNews

Tesla is collecting insane amount of data from its Full Self-Driving test fleet

Tesla is already collecting large amounts of data from the vehicles’ owners in its Full Self-Driving beta test fleet.

It is already releasing a new update improving on the release earlier this week.

Connectivity has been a flagship feature of Tesla’s offering in the auto industry.

It not only enables several fun and useful features for customers, but Tesla has also been using the connectivity to collect data from the fleet and improve features — mainly Autopilot.

When releasing an Autopilot update back in 2017, Tesla asked owners for the authorization to collect videos from the Autopilot cameras.

After that, Tesla opened the floodgates of Autopilot data gathering.

The data collection is voluntary for Tesla customers, but most owners are more than willing to let Tesla collect data since the automaker uses the data to feed its neural networks powering Autopilot and now Full Self-Driving systems.

Tesla can detect when drivers intervene during Autopilot operations and send back data to the mothership in order to help the neural networks improve.

This is also true now that Tesla has started rolling out its Full Self-Driving Beta update this week.

Tesla owners can monitor their WiFi connection to know when and how much data Tesla is pulling from their cars.

Right now, there are only a few Tesla owners in the limited release of the FSD Beta, but one of them monitored his car’s connection after a day of driving and managed to catch Tesla’s data collection.

He says that he saw Tesla pulling as much as 4 gigabytes of data from his car:

That’s a significant amount of data being pulled from the vehicle and it looks like Tesla is already pushing improvements less than a week after releasing the update to vehicles in its customer fleet.

Tesla has started pushing a new update (2020.40.8.11) to vehicles on the FSD Beta limited release.

Electrek’s Take

I wouldn’t be surprised if the limited release to customers didn’t even double the number of vehicles with the FSD build, but that’s still a lot more data to collect.

I think Tesla is going to use this current data collection effort with a small fleet to tune its data gathering ahead of opening the beta release to the wider fleet.

Much has been said about the amount of data Tesla is able to gather thanks to its large fleet of hundreds of thousands of vehicles equipped with sensors, but the quality of data is important.

Tesla needs to have an efficienct way to feed the relevant data to its neural nets for them to improve.

I think that’s exactly what we are seeing right now.


Subscribe to Electrek on YouTube for exclusive videos and subscribe to the podcast.


Author: Fred Lambert
Source: Electrek

Related posts
AI & RoboticsNews

Cohere just made it way easier for companies to create their own AI language models

AI & RoboticsNews

How GPT-4o defends your identity against AI-generated deepfakes

AI & RoboticsNews

Microsoft leans harder into AI, updating Copilot, Bing, and Windows

CryptoNews

Ripple Boosts Cross-Border Transactions in Brazil With New Payment Solution

Sign up for our Newsletter and
stay informed!