Rivian considering making own LiDAR sensors

[NY_Rob]

^ wouldn't be at all surprised... it was a "winner takes all" format war with hundreds of millions of dollars at stake. Sony had already lost the Betamax/VHS contest and had tons of $$$ invested in Betamax so if it took a little extra persuasion to win, so be it.

Sponsored

 

[therealcmj]

Including a Chinese made LiDAR sensor would effectively make every Rivian an espionage platform for the CCP

The LIDAR sensor just returns the point cloud to the computer it's attached to. There's no way for it to send data outside of the car by itself.

 

[mkg3]

It makes complete sense. The whole point is to control the whole stack and minimize inputs to the devices that they have no control over. RJ has said he wants to control the whole stack because that's how Apple has done business....

There is a huge difference between Apple and Rivian in terms of business profitability.

Apple only started to do their own silicon after years of being profitable. They bought P.A. Semi (silicon design firm) to acquire the domain knowledge. And it was only years after that they actually took the plunge to start their own silicon design.

Rivian is not profitable and continue to lose money and they have not acquired any domain expertise company who already design and make silicon.

What you say sounds logical and on the surface, it may make sense. Once one dives deeper into the company capabilities, and their positioning, at the current financial picture and their business actions to date, does not make sense.

 

[Zathras]

There is a huge difference between Apple and Rivian in terms of business profitability.

Apple only started to do their own silicon after years of being profitable. They bought P.A. Semi (silicon design firm) to acquire the domain knowledge. And it was only years after that they actually took the plunge to start their own silicon design.

Rivian is not profitable and continue to lose money and they have not acquired any domain expertise company who already design and make silicon.

What you say sounds logical and on the surface, it may make sense. Once one dives deeper into the company capabilities, and their positioning, at the current financial picture and their business actions to date, does not make sense.

You seem to have missed my point. Maybe not. But the point of having the whole stack in-house is they are not beholden to anyone else and can make their own mistakes, and move in a different direction without the nonsense of the politics of dealing with independent decision-makers.

Rivian seems to be doing really well with their vehicles considering the nature of their business up to this point. They are at the deflection point where Tesla was several years ago. The R2 is shaping up just fine, their self-driving system is growing fast. They have the right people in place. Who are they going to partner with to tell them where they are going wrong? Honda? Nope, they are in a massive deficit now because they're resetting their EV strategy that could take a decade to recover from. Toyota is just as bad off. Volkswagon believes in their strategy. How is this vote of confidence from the #2 car maker in the world not making sense? They cancelled the ID.4 to make this shift to adopting Rivian's electronics and software. Seems to me they are doing just fine.

With the R2 they should be in profitablity by 2028 or soon thereafter. What EV maker outside of China and Tesla say that? RJ is a smart guy, he seems to be hiring the right people. And he's doing what Apple couldn't (make a car) and the reason why is because they are not partnering with legacy car makers (other than VW and Uber buying in to their hardware/software stack) who have no clue what to do.

I could be wrong. I don't think so. (Self delusion? We'll see.)

 

[Zoidz]

There is a huge difference between Apple and Rivian in terms of business profitability.

Apple only started to do their own silicon after years of being profitable. They bought P.A. Semi (silicon design firm) to acquire the domain knowledge. And it was only years after that they actually took the plunge to start their own silicon design.

Rivian is not profitable and continue to lose money and they have not acquired any domain expertise company who already design and make silicon.

What you say sounds logical and on the surface, it may make sense. Once one dives deeper into the company capabilities, and their positioning, at the current financial picture and their business actions to date, does not make sense.

Apple was involved in silicon design before becoming profitable. In fact, Apple's investment in 1990 as an ARM founding partner did three things - it put ARM on the tech map, it established the foundation still used in mobile devices today (via the failed Newton) and helped to saved Apple's ass in the late 90s when they sold their investment in ARM for $1.1B.

In today's world, you don't need to aquire another company to design ASICs and contract fab is easy peasy at this level, you just need to hire engineers who know how to use the design software. Mukund Chavan, Vice President of ASIC Design, worked for Telsa and Apple prior to joining Rivian. The dude has the chops. Rivian's early investment in silicon design could turn out to pay off strategically in several possible ways, just like it did for Apple.

 

[kurtlikevonnegut]

The LIDAR sensor just returns the point cloud to the computer it's attached to. There's no way for it to send data outside of the car by itself.

My friend, finding ways to get sensors to send data to the mothership that they aren't supposed to be able to send is basically a full time hobby for the CCP. If they are able to do so with extremely simple devices I'm sure that they can manage with a highly complex machine.

 

[captainjp]

My friend, finding ways to get sensors to send data to the mothership that they aren't supposed to be able to send is basically a full time hobby for the CCP. If they are able to do so with extremely simple devices I'm sure that they can manage with a highly complex machine.

Well then should we all toss our iPhones with FaceID since they’re incorporated with LiDar from china?

 

[therealcmj]

My friend, finding ways to get sensors to send data to the mothership that they aren't supposed to be able to send is basically a full time hobby for the CCP. If they are able to do so with extremely simple devices I'm sure that they can manage with a highly complex machine.

If there's no radio in the LIDAR sensor there's literally no physical way it can send the data.

If they compromise the car's computer that is receiving the point cloud then there's no need to do anything in the LIDAR sensor.

But all of that is stupid when they could just compromise the servers that contain the data.

 

[Zoidz]

I really hope that they don't do this. Including a Chinese made LiDAR sensor would effectively make every Rivian an espionage platform for the CCP, and it would likely mean that any future Rivian vehicles would be out of consideration for me.

The Chinese are reported to be Rivian's LIDAR "off the shelf" manufacturer if they don't do their own design - Robosense. If Rivian does their own design, Rivian has MORE control over the hardware device.

But at any rate, LIDAR sensors are hardwired devices and it will be controlled by RIvian's software stack, which in turn only communicates to Rivian's cloud infrastructure. Short of Rivian being fully compromised, exactly how do you think China will get access to the LIDAR sensor? And what value is it to them? There is other more valuable data in the cloud that is easily accessible by China.

Time to take off that tinfoil face mask and get a fresh breath of air.

 

[mkg3]

You seem to have missed my point. Maybe not. But the point of having the whole stack in-house is they are not beholden to anyone else and can make their own mistakes, and move in a different direction without the nonsense of the politics of dealing with independent decision-makers.

Rivian seems to be doing really well with their vehicles considering the nature of their business up to this point. They are at the deflection point where Tesla was several years ago. The R2 is shaping up just fine, their self-driving system is growing fast. They have the right people in place. Who are they going to partner with to tell them where they are going wrong? Honda? Nope, they are in a massive deficit now because they're resetting their EV strategy that could take a decade to recover from. Toyota is just as bad off. Volkswagon believes in their strategy. How is this vote of confidence from the #2 car maker in the world not making sense? They cancelled the ID.4 to make this shift to adopting Rivian's electronics and software. Seems to me they are doing just fine.

With the R2 they should be in profitablity by 2028 or soon thereafter. What EV maker outside of China and Tesla say that? RJ is a smart guy, he seems to be hiring the right people. And he's doing what Apple couldn't (make a car) and the reason why is because they are not partnering with legacy car makers (other than VW and Uber buying in to their hardware/software stack) who have no clue what to do.

I could be wrong. I don't think so. (Self delusion? We'll see.)

Did not miss the point about the vertical integration of Rivian technology use. My point is that they are not in good position to do so currently. This is where I think RJ the engineer comes out over RJ the business man.

As for Rivian being profitable in 2028, that's not fast enough when R2 is being derived soon. And will they be profitable per vehicle they make, rather than all the other income sources.

Now retired but have spent my entire career in aerospace industry leading programs with complex integration of systems, the make-or-buy decision is always made with P&L responsibility. As such, the question goes back to is now the right time? From my perspective, using one of the COTS LiDAR and learning about it more and its strengths and weaknesses would allow Rivian to learn and be more cost effective, if they so choose to build their own.

 

[mkg3]

Apple was involved in silicon design before becoming profitable. In fact, Apple's investment in 1990 as an ARM founding partner did three things - it put ARM on the tech map, it established the foundation still used in mobile devices today (via the failed Newton) and helped to saved Apple's ass in the late 90s when they sold their investment in ARM for $1.1B.

In today's world, you don't need to aquire another company to design ASICs and contract fab is easy peasy at this level, you just need to hire engineers who know how to use the design software. Mukund Chavan, Vice President of ASIC Design, worked for Telsa and Apple prior to joining Rivian. The dude has the chops. Rivian's early investment in silicon design could turn out to pay off strategically in several possible ways, just like it did for Apple.

Yes they developed PowerPC RISC chip with IBM, which almost tanked the company and was not successful at most of the chips they made. The original iPod processor was designed and PortalPlayer, which apple funded the development. After Jobs came back and decided to use Intel x86 processors for Macs made save the entire computing line.

ARM license the core design that others build on. It has taken many years for Apple to bring in most of the silicon that goes into iPhone - cell modem being the most recent.

In today's world, as you say, with all the AI help, it is imperative that one possess domain knowledge to rapidly implement poducts without much trail and error and without a steep learning curve. Rivian has to make their ADAS chip work and produce successfully in mass, I believe, before taking on another chip design on their own.

 

[Zathras]

Yes they developed PowerPC RISC chip with IBM, which almost tanked the company and was not successful at most of the chips they made. The original iPod processor was designed and PortalPlayer, which apple funded the development. After Jobs came back and decided to use Intel x86 processors for Macs made save the entire computing line.

ARM license the core design that others build on. It has taken many years for Apple to bring in most of the silicon that goes into iPhone - cell modem being the most recent.

In today's world, as you say, with all the AI help, it is imperative that one possess domain knowledge to rapidly implement poducts without much trail and error and without a steep learning curve. Rivian has to make their ADAS chip work and produce successfully in mass, I believe, before taking on another chip design on their own.

I think their profitability being a few years away is just fine. How long did they go before being profitable. And when they scale up to 400,000 units a year, we'll see if they made the right choice. It's not like the competition is an example of success in this area by any means. Tesla being the sole exception.

Apple finally decided to do it right by licensing the ARM tech and then built their own processors with the best minds in the business at the time. Apple has made many mistakes, and continues to do so. I just remember Microsoft and Blackberry laughing at them for doing it the way Rivian has been going.

 

[maba2366]

The most likely scenario is that they will license tech from a chinese company and build the chips domestically with their own software stack. This alleviates concerns about Chinese spy software in their cars. They were probably planning on using Luminar chips but after Luminar unraveled they need a plan B.

 

[Zathras]

The most likely scenario is that they will license tech from a chinese company and build the chips domestically with their own software stack. This alleviates concerns about Chinese spy software in their cars. They were probably planning on using Luminar chips but after Luminar unraveled they need a plan B.

No doubt. They have someone else make the matrix headlights, but they write the software. So it can be a combination of things. But they will control the headlines, and LIDAR, and anything else that they put in the car as a cohesive whole. They aren't going to give up control of the core system.

Sponsored