Home » Electric Cars » “It can see dogs!” The big reveal from Tesla’s radar-free version of full self-driving

“It can see dogs!” The big reveal from Tesla’s radar-free version of full self-driving

fsd 9 dogs
Source: Twitter

Tesla has released the hotly anticipated latest version of its Full Self-Driving software suite (FSD v9.0) and it claims to be able to do things such as spotting dogs by the side of the road to handling intricate intersections. But will it be able to spot kangaroos?

The new v9.0 software rollout to the small number of FSD beta testers that have been testing Tesla’s fully autonomous driving package since late 2020 was a little behind schedule, having been touted by Tesla boss Elon Musk for a mid-May launch, but its reception has certainly been positive.

And that’s just as well, having been hyped somewhat by Musk who has variously said the v9.0 release would be “mind-blowing” and “gigantic“, and that the new FSD would feature a new “mind of car” visualisation in the new release.

It is the first version of Tesla’s FSD package to be released without the use of radar sensors to detect objects. It now only uses the suite of cameras embedded in various positions all around a Tesla car’s body.

New images emerging on Twitter under the hashtag #FSDBeta9 give us insight into what the new FSD release is capable of.

It’s not so much mind-blowing as just visibly safer. If you’ve been watching beta testers uploading their content on Youtube and Twitter, you’ll know that the earlier versions of FSD beta relied on human intervention to learn how to handle difficult road situations such as roundabouts.

The learning curve has taken longer than an optimistic Musk expected, leading him to admit that “the difficult is obvious in retrospect.”

But it’s getting there, as these examples demonstrate. The instance below shows how one FSD beta tester in a crowded beach carpark automatically stops for random pedestrians:

Mind you, that is at a very slow speed. But we are also beginning to see tricker manoeuvres being handled such as this right-hand turn and immediate lane change.

Another example shows pedestrians being spotted between and behind vehicles. Whether this is because they have been seen through a window or extrapolated from the direction they were walking before going behind a vehicle is not clear.

Interestingly, the FSD visualisation now uses more realistic visuals akin to the current FSD but it would seem it can recognise more types of objects, such as dogs.

As Kim Paquette points out, the previous version rendered dogs as small boxes.

https://twitter.com/kimpaquette/status/1413898519086342144

It’s details like this, and the ability to stop for them, that will mean Tesla’s autonomous software is more likely to be approved by regulators for widespread use.

But it also adds to the fascination, at least for this writer. That a couple of cameras can spot, recognise and then visualise a dog by the side of the road is smart.

But here in Australia we need it to spot kangaroos – when it can do that we’ll consider it ready for the outback!

Bridie Schmidt is associate editor for The Driven, sister site of Renew Economy. She has been writing about electric vehicles since 2018, and has a keen interest in the role that zero-emissions transport has to play in sustainability. She has participated in podcasts such as Download This Show with Marc Fennell and Shirtloads of Science with Karl Kruszelnicki and is co-organiser of the Northern Rivers Electric Vehicle Forum. Bridie also owns a Tesla Model Y and has it available for hire on evee.com.au.