Rivian considering making own LiDAR sensors

[savethemanual]

In house Silicon, now LiDAR?!?

https://www.reuters.com/business/au...possibly-partnership-with-chinese-2026-05-06/

• Rivian may partner with Chinese firms to develop lidar sensors for self-driving tech
• CEO cites Chinese dominance in affordable lidar tech
• Rivian investing hundreds of millions in custom chip program
• Will make at least two successors to first chip

SAN FRANCISCO, May 5 (Reuters) - Rivian Automotive is considering making its own lidar sensors and ‌could do so in partnership with a Chinese firm, CEO RJ Scaringe said in an interview with Reuters on Tuesday.

The Irvine, California-based electric automaker last year launched a plan to make its own chips as part of a broader push to develop its own proprietary self-driving technology that would compete with products from Tesla (TSLA.O), opens new tab. Rivian said that a version of its R2 vehicles coming later this year would include lidar sensors, which help self-driving vehicles gain a three-dimensional view of the road.

Sponsored

 

[savethemanual]

Also, here's a photo from his recent interview with Reuters....now we know the true inspiration for Borealis

 

[NY_Rob]

Interesting!

One EV manufacturer trying to find a way to include LiDAR in it's vehicles and another has gone out of it's way to not use it.

 

[captainjp]

Blu-Ray vs HD DVD.

 

[mkg3]

This makes no sense. There already are low cost LIDAR chips and making chips is not Rivian's core competency.

It has potential of being rabbit hole and materially impact earnings. This is where RJ need to be s smart make-or-buy executive.....

 

[Davethadog]

Blu-Ray vs HD DVD.

Two things that were seen as equivalent at the time but one was technically superior.

 

[NY_Rob]

Two things that were seen as equivalent at the time but one was technically superior.

Yeah, and the superior format was not the one which survived unfortunately.

 

[iamnid]

Interesting!

One EV manufacturer trying to find a way to include LiDAR in it's vehicles and another has gone out of it's way to not use it.

Mercedes also recently announced it's dropping LiDAR and going to vision only.

 

[Mos Eisley]

This makes no sense. There already are low cost LIDAR chips and making chips is not Rivian's core competency.

It has potential of being rabbit hole and materially impact earnings. This is where RJ need to be s smart make-or-buy executive.....

It makes absolute sense. They are driven to make Rivian’s 100% vertically integrated and they have the backing of Amazon to make this happen through their last-mile fleets. Owning most of your supply chain top to bottom can make you more nimble and efficient. After RJ’s comments about how many manufacturers across three tiers of supply chains (it was more than 50!) just to make the headlight… I can understand why they’d have motivation to own more, more directly.
And while this news doesn’t imply that current timelines for R2 LiDAR might be stretched, I could see it happening. What RJ mentioned here is probably more likely the 2nd or even 3rd gen of lidar.

 

[Mos Eisley]

Relevant:

 

[NY_Rob]

Mercedes also recently announced it's dropping LiDAR and going to vision only.

Like Tesla, time will tell if they took the right approach.

 

[ElGuano]

It makes absolute sense. They are driven to make Rivian’s 100% vertically integrated

Hard agree. It seems easy to say "just use an available 3p solution, but we've seen over and over that this can present present and future limitations that significantly impact plans, some major, some minor.

1. Tesla and Rivian started camera autonomy with MobileEye. This was eventually found to be limiting for both, and both moved to internal hardware/software stacks.

2. Tesla originally relied on 3p maps APIs for navigation. But it was laggy, didn't allow for fine tuning the visualizations or other features such as dynamic waypoints, etc. Around 2019 they rewrote it from scratch, and the difference was dramatic (framerate doubled, nav accuracy improved, etc.)

3. Rivian relied on 3p hardware controllers for seat adjustments in the gen1 R1. In Gen2 they moved to an internal stack with the change to zonal architecture, and now we can ... move and tilt the seats simultaneously!

I'm sure there are available and affordable LIDAR sensors out there today. But do they present a (oof) "supply chain risk?" Do they allow for the low level access Rivian would want to fine-tune autonomy in the conditions needed (snow, rain, etc.)? Does the 3p roadmap align with what Rivian will need in 12 months? 36 months? Vertical integration is a serious consideration for something like a car.

 

[Ralph]

This makes no sense. There already are low cost LIDAR chips and making chips is not Rivian's core competency.

It has potential of being rabbit hole and materially impact earnings. This is where RJ need to be s smart make-or-buy executive.....

Not sure I agree with your first statement. The balance between strategic vertical integration or being a slave to "core competency" involves inputs that we on the outside can't be fully privy to for obvious reasons. Most of us couldn't understand the interrelationships if we were aware of them.

This is not necessarily inconsistent with your second statement.

 

[godfodder0901]

Also, here's a photo from his recent interview with Reuters....now we know the true inspiration for Borealis

More like Borealis was the inspiration for the shirt:
Shepherds Shirt

 

[DuoRivians]

If Rivian is still trying to figure this out, it doesn’t speak well to when we’ll actually see lidar in R2

Sponsored