back
23 comments
As mentioned at the end of the article, there are other applicants than just self driving. Body scanning can be one that can be useful for building 3D apps.
>Body scanning can be one that can be useful for building 3D apps

Oh boy, can't wait till Amazon, Meta and Google can monetize my body and insecurities. /s

Well, it's probably not yet monetized but they almost certainly have the data. Dimensions like gait, gaze, stance, and a hundred other more subtle measurements are being collected via their VR headsets at various levels of anonymity and granularity.
They also just know whatever you have ever purchased, including what size jeans you bought.
Haven’t read the article yet, but there’s a lot of old mines in my area and I’d love to build a self exploring drone to map these mines using LiDAR.
It's not a problem today. Check online, there are cheap small ToF lidars. Like the one I bought with 60 degree, 100x100, 15 f/sec. 8 of them will give you all around horizontal view, up and down. 3m range should be enough. Your main problems will be lack of GPS and short flight time.
Forgot to say, the complexity of this, if you manage to do it, is a solid PhD. At least one.
oh hey! a goal to aim for.
Before you know it, every kebab cart in NYC will be selling LIDAR on a stick
Bubba Gump's LIDAR Company for autonomous shrimpin' boat guidance
I thought the Apple iPhones had a LiDAR on a chip? https://www.edge-ai-vision.com/2020/05/with-the-apple-ipad-l...
Maybe the difference is that chip is ToF-style (critiqued in the article), while this new chip will be FMCW
Cheap, effective, reliable, durable lidar in cars would be a good thing.

But for self-stopping, not self-driving.

Reminds me of the ancient volvo(?) commerical, where instead of stopping for an obstacle, it drives around and avoids it.
This is really neat, there are some really cool, low cost integrated packages for ToF sensors (not what the article was talking about though). ST makes a bunch of them and they are surprisingly small.
How does it work when there are millions of these chips out there. I'm not an EE but is the interference a trivial probablem at this point?
they talk about this in the article, apparently there's an analogy to AM vs FM radio to be made with "FM" much more tolerant of interference in lidar
Maybe Tesla will change their tune on LiDAR as a result
I'd probably buy Tesla stock if I thought they were going to announce this switch. It would mean real progress instead of what ever hell they're currently stuck in.
The problem with that idea is that if Tesla ever gets out of fantasyland, their valuation is absurd, and the stock would go down regardless.
Ya probably hah
Would be interested to hear how this capability at a low price point will affect other industries beyond self-driving cars...
The working prototype is pretty honking for a lidar on a chip