Tesla bought over $2M worth of Lidar from Luminar

In order to make vision-only depth sensing work well, you need ground-truth data. Buying a few thousand lidar to get a ton of ground-truth training data for the NN to learn distance on its own from vision is an obvious move. They're not switching to Lidar.

source: i debugged lidar and built localization and behavioral planning algorithms for self driving cars

Re: Tesla bought over $2M worth of Lidar from Luminar

#7

Post by grecy » Wed, May 08, 2024, 5:12 PM UTC

lenerdenator wrote: post #5

How much Lidar does $2 million buy you? Probably, what, a year's worth of Tesla's manufacturing output?

Not even remotely close.

This is clearly R&D just to validate the images from their vision approach.

Re: Tesla bought over $2M worth of Lidar from Luminar

#8

Post by cozzyd » Wed, May 08, 2024, 5:13 PM UTC

lenerdenator wrote: post #5

How much Lidar does $2 million buy you? Probably, what, a year's worth of Tesla's manufacturing output?

enough for one engineer's test bench, more like

Re: Tesla bought over $2M worth of Lidar from Luminar

#9

Post by sushid » Wed, May 08, 2024, 5:15 PM UTC

ckdarby wrote: post #3

Very confused to why they're not just acquiring. If they're representing 10% of revenue are they're doing a pilot and then considering outright buying them?

100% revenue != valuation of company if that's what you're implying.

Re: Tesla bought over $2M worth of Lidar from Luminar

#10

Post by stetrain » Wed, May 08, 2024, 5:17 PM UTC

lenerdenator wrote: post #5

How much Lidar does $2 million buy you? Probably, what, a year's worth of Tesla's manufacturing output?

They produced 1.85 million vehicles in 2023.

So a $2 million order would work out to $1.08 per car.

Or if Lidar hardware for one car cost about $500, then $2 million is not quite enough for one day's worth of manufacturing output.