A "Play by Play Illustrated Commentary" why R2 production would be challenging to retrofit into existing R1 or EDV line

[Zoidz]

There have been a number of threads about Rivian should do this, that and the other to get R2 production up as quickly as possible, such as converting one existing line to another, sharing lines, etc. I'm no expert on automotive lines, but I have been down this road on other types of manufacturing and production line.

My speculation is that the R2 will need to be a new line due to thousands of challenges presented by trying to convert or share an existing line and the time it would take to solve all these issues AND not affect existing/ongoing production schedules of R1 or EDV. It should be possible to share some of the work cells that already have flexible operations capability such as stamping and painting. To help visualize these engineering challenges, here are some screen shots of a video tour with RJ of the Normal factory. I encourage everyone to watch the video. It should help in understanding just how challenging any retooling or sharing will be. Any one challenge may be"easy". But multiply that times 1000, and then integrate them to all work together. Not so easy.

Stamping: Since the R2 is smaller, the existing stamping shop could in theory be retooled using the same machines. It would require new dies and motion control ("robotics") re-programming. Retooling is a normal operation in most stamping shops, so IMO it's just time to get dies made and prove it in. It's probably one of the easier processes to update for R2. But storage racks would need to be new. In the third picture below, this is probably an R1T side panel. Note how the rack is design to hold R1T panels with a support through the tunnel. That does not exist on the R2S, and R2S panels are smaller than R1S. So new racks would be needed. But in turn, that ALSO affects the robotic operations on the body shop weld line - storage racks are different, panel sizes are different..... so what has to be reworked on the existing body shop weld line?

Body Assembly Shop:
Look at the size of the R1 panels and the robots, and how closely the robots are spaced together. In the second picture, note the number of jig fixture attachment points. All that is specific to the R1. This all needs to be redesigned for the R2, and THEN figure out how to make it a flexible changeover process between R1 and R2 builds, if you built both on the same line.

Now look at the EDV body shop. These panels and robots are MUCH larger and it looks to me like this is a fixed work cell process, not a continuous moving assembly line process. RJ comments that they look like dinosaurs. There's just no way EDV could be built on the R1 line.

Paint Shop:
RJ says in the video that the paint shop already processes the R1 and EDV, and also does the skateboard electrostatic protection coating. The paint shop could be reprogrammed fairly easily to add the R2. The problem is it probably is a production bottleneck, so factory throughput of all products would likely be constrained by this work cell. In the third and several other pictures, notice the jigs that the R1T sits on during various processes. This jig (Rivian calls it the "R1 Skillet Skid" in final assembly) is designed to the R1 vehicle, so an inventory of hundreds of new R2 jigs and skillet skids will be needed. And again, every workcell may need upgrades, updates, modifications, programming due to the different size of the jig/skid.

General Assembly
Notice in the following pictures how all of the jigs are made to R1 dimensions. What would it take to retool and reprogram everything you see in these pictures to adjust for dimensional differences of R2 or EDV on the same line?

Sponsored

 

[COdogman]

Great insights. Thanks for posting this.

The scale of these manufacturing processes are hard to imagine until you see it in person.

I do hope they can get the R2 to market ASAP but I’m not holding my breath.

 

[Dark-Fx]

It might be possible to share hard points across the R1S/R2 models, but I'm doubtful considering how much skinnier R2 is. Maybe there is enough difference to nest the R2 hard points inside the R1 platform, but they would have had to design R2 with this stuff in mind, and all of the jigs too.

I wouldn't be surprised to find out that the reason it took them this long to announce their pivot is because they wanted to solve all of the engineering challenges of doing this first. Which should mean they are in a lot better position to do it than we think.

 

[Mark_AZR1T]

It will absolutely be a new line in
Normal, which simply leverages all resources in house. Apparently, they've applied for additional state of Illinois assistance in financially doing this, so it makes sense to build a new line.

First deliveries Q4 2025, this would be a massive under promise over deliver for the team at Rivian, but really their existence as a entity on their own rests on it

 

[defcon888]

There have been a number of threads about Rivian should do this, that and the other to get R2 production up as quickly as possible, such as converting one existing line to another, sharing lines, etc. I'm no expert on automotive lines, but I have been down this road on other types of manufacturing and production line.

My speculation is that the R2 will need to be a new line due to thousands of challenges presented by trying to convert or share an existing line and the time it would take to solve all these issues AND not affect existing/ongoing production schedules of R1 or EDV. It should be possible to share some of the work cells that already have flexible operations capability such as stamping and painting. To help visualize these engineering challenges, here are some screen shots of a video tour with RJ of the Normal factory. I encourage everyone to watch the video. It should help in understanding just how challenging any retooling or sharing will be. Any one challenge may be"easy". But multiply that times 1000, and then integrate them to all work together. Not so easy.

Stamping: Since the R2 is smaller, the existing stamping shop could in theory be retooled using the same machines. It would require new dies and motion control ("robotics") re-programming. Retooling is a normal operation in most stamping shops, so IMO it's just time to get dies made and prove it in. It's probably one of the easier processes to update for R2. But storage racks would need to be new. In the third picture below, this is probably an R1T side panel. Note how the rack is design to hold R1T panels with a support through the tunnel. That does not exist on the R2S, and R2S panels are smaller than R1S. So new racks would be needed. But in turn, that ALSO affects the robotic operations on the body shop weld line - storage racks are different, panel sizes are different..... so what has to be reworked on the existing body shop weld line?

Body Assembly Shop:
Look at the size of the R1 panels and the robots, and how closely the robots are spaced together. In the second picture, note the number of jig fixture attachment points. All that is specific to the R1. This all needs to be redesigned for the R2, and THEN figure out how to make it a flexible changeover process between R1 and R2 builds, if you built both on the same line.

Now look at the EDV body shop. These panels and robots are MUCH larger and it looks to me like this is a fixed work cell process, not a continuous moving assembly line process. RJ comments that they look like dinosaurs. There's just no way EDV could be built on the R1 line.

Paint Shop:
RJ says in the video that the paint shop already processes the R1 and EDV, and also does the skateboard electrostatic protection coating. The paint shop could be reprogrammed fairly easily to add the R2. The problem is it probably is a production bottleneck, so factory throughput of all products would likely be constrained by this work cell. In the third and several other pictures, notice the jigs that the R1T sits on during various processes. This jig (Rivian calls it the "R1 Skillet Skid" in final assembly) is designed to the R1 vehicle, so an inventory of hundreds of new R2 jigs and skillet skids will be needed. And again, every workcell may need upgrades, updates, modifications, programming due to the different size of the jig/skid.

General Assembly
Notice in the following pictures how all of the jigs are made to R1 dimensions. What would it take to retool and reprogram everything you see in these pictures to adjust for dimensional differences of R2 or EDV on the same line?

Awesome and thorough post.

Here is my 2 cents.

Rivian bought the old Mitsubishi plant and from what I read it was able to product 225,000 vehicles a year at it. You and I both know that Rivian isn't even close to doing 100,000 let alone 225,000/year.

There more than likely is enough space to put in machinery, robots, stamping, paint, electricaletc. I am like you, not an expert in automotive production but knowing a little about production, I can see something like that happening.....at least for the initial run of R2S. They could then continue building the plant in Georgia and eventually populate it with machinery or just shut down for a month and move the equipment there...might be more of a Cost.

This is an aerial view of Rivian in Normal. Maybe they can build where the rectangle is and buy the land where the triangle is and move the parking lot and service center over there and expand into the parking lot.

 

[DuoRivian]

The existing factory was expanded to 200k of which 100k covers R1 and EDV. Makes sense to have a new line and use the existing paint shop which should have the capacity for 100k R2’s. They have to get this up and running in 18 months to allow early builds before end of 2025. I don’t see them building major buildings in that time.
As said earlier they have probably worked on this option for a few months and have solutions.

What I wonder about is when Georgia will restart because that will take 18-24 months to finish and is needed for additional R2 capacity (North America and Europe) and R3.

I would expect a lot from the Georgia plant like stamping and general assembly jigs can be used at Normal. Use the existing battery and paint shop facilities and build a new general assembly and stamping/assembly line nearby.

 

[defcon888]

The existing factory was expanded to 200k of which 100k covers R1 and EDV. Makes sense to have a new line and use the existing paint shop which should have the capacity for 100k R2’s. They have to get this up and running in 18 months to allow early builds before end of 2025. I don’t see them building major buildings in that time.
As said earlier they have probably worked on this option for a few months and have solutions.

What I wonder about is when Georgia will restart because that will take 18-24 months to finish and is needed for additional R2 capacity (North America and Europe) and R3.

I think once the R2s start rolling off the line and the cash starts flowing in, they can see the demand and project the needs. This will help kickstart continuing building.

RJ is a smart guy, I think they have contingency plans for the contingency plans.

I believe in Rivian, so much I signed up for a brokerage account and bought a few shares of RIVN.

 

[KootenayEV]

Awesome and thorough post.

Here is my 2 cents.

Rivian bought the old Mitsubishi plant and from what I read it was able to product 225,000 vehicles a year at it. You and I both know that Rivian isn't even close to doing 100,000 let alone 225,000/year.

There more than likely is enough space to put in machinery, robots, stamping, paint, electricaletc. I am like you, not an expert in automotive production but knowing a little about production, I can see something like that happening.....at least for the initial run of R2S. They could then continue building the plant in Georgia and eventually populate it with machinery or just shut down for a month and move the equipment there...might be more of a Cost.

This is an aerial view of Rivian in Normal. Maybe they can build where the rectangle is and buy the land where the triangle is and move the parking lot and service center over there and expand into the parking lot.

One thing to keep in mind is that (as I understand it) Rivian is more vertically integrated and building more of their own parts than Mitsu likely was, so they need factory area for that, not just General Assembly for vehicle lines. They've already expanded the factory once IIRC to bring electric motors in-house.

 

[mikef]

I looked at the county property map at the parcels around the factory and Rivian already owns these fields west of it:

Also they've expanded the factory to the SW quite a bit the last year or two. They've taken up a parking lot and expanded the building toward the road to the W and S. That's the part of the video where RJ says something like "where I'm standing used to be the wall" or "is all new" or something like that. I think they build motors or batteries in that area now.

It was in the local news a few months ago Rivian bought or leased a bunch of warehouse space in the industrial park in the northern part of town NE of the interstate interchange.

I agree on the existing R1 line not being able to switch back and forth between R1 and R2. I've been on the tour. The first issue to come to mind is the way the frame of it is put together. The way it is now think of a table-frame where a human or robot drops the A pillar, B pillar, and other pieces into a frame then the weld bot welds the pieces together to form one side wall of the frame of the vehicle. They'd need a different frame for R2 to drop the different R2 pieces into.

The existing factory has 2 production lines, one for the EDV and the other for the R1 line. Here's the plan for the Georgia site which shows what activities are where in their process.

 

[HaveBlue]

R1 dimensions are not consistent. The two models have different dimensions and parts that go into them even in the same model but the same robots can be programmed to handle both. The same robots will be assigned to build new cars. It just takes a lot of recalibration. For instance a welding robot has to be overseen by a certified human welder for calibration and quality control. Many of the assemblies like gear boxes are assembled by hand as well. The transition from the R1S to the R2S will be the same as the R1T to R1S. Some parts will swap over and many won't.

 

[defcon888]

One thing to keep in mind is that (as I understand it) Rivian is more vertically integrated and building more of their own parts than Mitsu likely was, so they need factory area for that, not just General Assembly for vehicle lines. They've already expanded the factory once IIRC to bring electric motors in-house.

great point and I didn't know that

 

[Zoidz]

R1 dimensions are not consistent. The two models have different dimensions and parts that go into them even in the same model but the same robots can be programmed to handle both. The same robots will be assigned to build new cars. It just takes a lot of recalibration. For instance a welding robot has to be overseen by a certified human welder for calibration and quality control. Many of the assemblies like gear boxes are assembled by hand as well. The transition from the R1S to the R2S will be the same as the R1T to R1S. Some parts will swap over and many won't.

Most of the existing robots could probably be reprogrammed for R2 on a new line. But unless you work for Rivian or an OEM supplier as an engineer, you simply can't make a blanket statement about what it will take to manufacture R2 in Normal.

 

[HaveBlue]

I don't think I made a blanket statement. Just a speculation like you have. The R1T and R1S already vary quite a bit where panel sizes, number and type of doors, interior fitment, frame etc. Even within the same model there are motor, battery, skid plates and things that vary. I'm not an automotive engineer either but also in a production oriented business. I have built hundreds of homes as a developer, as many as 60 a year and definitely a production process is complicated with lots of moving parts. Normal is still the best choice to get production going. The weather has definitely delayed things and burned time.

 

[sub]

unless you work for Rivian or an OEM supplier as an engineer, you simply can't make a blanket statement about what it will take to manufacture R2 in Normal.

And an extremely high level Rivian employee (RJ) said that they can manufacture it in Normal. So it sounds like you both agree that it can be done in Normal.

 

[Zoidz]

And an extremely high level Rivian employee (RJ) said that they can manufacture it in Normal. So it sounds like you both agree that it can be done in Normal.

It certainly can be done in Normal on a new line. But I think it is not viable on an existing line without completely disrupting current production.

Sponsored