The article unfortunately doesn't mention how the new method saves paint. I think the traditional electrostatic method has low losses already.
I've been in the BMW paint shop in their Munich plant two times. What I remember is paint going to the vehicle body - and only to the vehicle body. There is no mess at all, the room stays super clean and it has to, because subsequent cars are painted in different colors. There is no extra clean-up when changing colors. It would be interesting to know how the new Inkjet method stacks up against this.
One area where I see an advantage is metallic paint. From what I understand, metallic paint isn't applied electrostatically and the losses are higher there.
Without any in-depth knowledge into painting - my understanding from the article is that this is mostly about two-tone (or more) painting. When you use a masking tape, you overspray and some of it goes onto the tape. If it's an intricate pattern, there's a lost of masking going on. With this method, you reduce/remove the need for masking and spraying over the tape, therefore reducing paint waste.
The main money saving yes but there are other savings at play which probably have fairly minimal $ cost but still matter.
Like the cost of running the compressed air system, extra clean room filtration etc. Reduction or elimination of these ancillary systems probably also factor into power savings.
All in all, just because it might not be cheaper doesn't mean that it's not net positive in other areas.
It will be interesting to see how the paint for this holds up. One thing you'll notice is as companies try to switch from oil to water based paints the bases just don't last. There is also this trend to find ways to put less coats of base and clear down for cost savings. That obviously doesn't last as long either. So whatever they need to run this through the head I have to question if people will be happy with it over time compared to something simple like a sticker.
Many, many Mercedes from the late 90s and early 2000s were sent to an early grave from their absolutely disastrous water-based paint, done in the name of being environmentally friendly. (Mine was one of them after several also not eco-friendly trips to the body shop along the way.)
It’s not particularly environmentally friendly to have to make 10 million extra cars to replace those which rusted out prematurely.
That was also the era of Mercedes-Benz using biodegradable insulation on their wiring harnesses and it resulted in all sorts of cars having to be scrapped long before they broke down mechanically.
I'm guessing the harness of the paint was the issue - virtually all of them would start to rust from the beltline door trim, presumably because the trim itself managed to wear through the paint.
Any resident Germans may feel free to correct me, but it's my understanding that, in Germany, any rust at all on a vehicle will disqualify the vehicle during annual safety inspection. This is why German cars are typically known for having the longest last paint jobs.
> It will be interesting to see how the paint for this
> holds up. One thing you'll notice is as companies try to
> switch from oil to water based paints the bases just don't last.
The process described in the fine article is for painting e.g. stripes, logos, and other traditional "decals". It is not for painting the chassis, unibody, underbody, or even the finish on most of the exterior. That is all still done by traditional methods.
Its also used for whole roof with two-tone cars, where roof is different color than the rest of the body. You save masking process and whole another run trough paint process.
The current water based paints BMW uses are terrible for impact resistance/rock chips (compare the front bumpers of pre-2012/2013 to cars manufactured after that era) but generally very good for UV stability and being able to correct defects/swirls.
Is this a new technology or just a new use case for an existing process? Inkjet print heads are very close to the surface being painted but this process appears to have a few inches of clearance.
I suspect the paint has additives to help with this process. We didn’t see any color other than black in this demo.
The video doesn’t show any close up pictures of the process but it looks novel.
I've seen this in UV printers, too (UV = inkjet, but the ink is cured with UV instead of solvent evaporation).
I poked around to figure out how they can print from so far away, but couldn't find any sources talking about it, and if printing from ~10cm away needs special tricks to accomplish or not.
The one trick I can think of to extend print distance would be increasing the drop size. That would probably be acceptable when printing on cars (the article) or mugs (UV printers) -- is that the trick?
Industrial inkjet designs can print from a very large distance, what I think is new on this one is using a robot arm to be able to print on different contoured surfaces.
Cool concept, looks super slow relative to exciting assembly plant paint though. Upside is you can buy a plaid car from the factory, downside is it’s +$5k and an extra month lead time.
Considering there's already 6 month waits on new cars, I'd happily throw in an extra month if it meant I got to use an online editor to design my own paint job.
I think that is relative to custom paint vs existing manual methods. That i would agree with, i just don’t see this replacing single tone paint methods. Which likely wasn't why it was built, so not sure why i'm pointing that out haha.
It saves money in air compressor energy costs, but the article glosses over how long this takes compared to traditional methods. If it takes 10 times as long to paint a car this way then you have to replicate it 10 times in the factory to maintain the same rate of production which could very easily end up costing more.
On the other hand, masking is a pain in the ass and takes a lot of time so this could end up being faster all around so long as you're comparing with complex multi-tone paintjobs.
Off the top of my head: tailors embroider the clothing they make for customers and luxury clothing companies like Luis Vuitton engrave the customer's name or a message on products like wallets. IIRC even Apple used to engrave custom messages on some of their products for free.
Custom embroiders/engraves are basically a way to reduce the second hand market as fewer people want to buy a luxury product with someone else's name on it.
I would think that if they were really trying to check the secondary market, they’d allow engraving for iPhones. Surely that’s by far the one that cuts most deeply into the bottom line.
Looking at the orange base-coat in the video, I'd suggest it saves money versus hand-painted custom paint jobs but as every surface of the car (inside doors) are orange, I'd say that was applied by the traditional robot painters (with HVLP guns).
The orange car in the picture seems to have already been dipped in coating and maybe "baked", and then painted with orange "aerosolized" paint. The inside of the car has orange overspray all over, and where it is thin we see the brighter yellow undercoat.
Basically, this is a way to print graphics on an already painted car.
It's a shame the article focuses so heavily on the green aspect rather than the process, but I guess BMW doesn't want to give away trade secrets, and this is just a regurgitated press release.
For example, what's going on in the first stage of the process where the paint head seems to be drawing but nothing shows up. It's it painting liquid masking tape onto the car? Just thin lines of the same paint? The robot appears to switche paint heads after doing it, suggesting it's maybe a different substance, or maybe just a different spray head.
The head have a laser measurement tool and measures the exact position of the motor hood with sub milimeter precision. Then calculates corrections for its program to apply at the exact spot everytime.
Disclaimer: I work robots in paintshops like these.
Example:
The inside of your car door have plastizol sprayed over whole edge. That is actually sprayed while the doors are (almost) closed (think open as if latched by the lock, but not flush). The nozzle is a 0.5mm wide tube, that makes 3 90° degrees turns (left, right, right), and have opening in the direction pointing back at where it started. When the carbody gets into position, fixed cameras measure preset points to make sure its in right position (wiggle room ~20-100mm). Correction gets computed from that, and robot measure pass program is moved accordingly. Now the measure process is repeated with robot head holding a laser tool, measuring tenths of milimeter deviations, wiggle room ~2mm.
If the shape is just moved within tolerances, the robot modifies apply run with corrections from measurement and goes to work. Starting at corner, weaves the nozzle into position and runs along the whole doors (sometimes pulling out and back in at some corners, but some corners are turned without pulling out). Theres usually at least half a millimeter spacing between nozzle and an edge, the precision is maintained even during fast movements.
The nozzle is thin and long (5 cm) and easily breaks or bends.
Before every spray run the tool itself is measured (and the correction is computed and used), and used only if its not bent out of tolerance. If the tool is found faulty, robot can hotplug another nozzle (have like 5 ready), measure it and go fully autonomously.
Theoretically, you dont need any CAD. The robot doesnt care about actual surface geometry. In fact, you can just move robot axis by axis (or by xyz) "by hand" (looks like this [0]), and teach points as you go.
Usually CAD is provided ahead and you can "teach" the points virtually on PC, simulate it and everything, before testing it with real robot and carbody.
For exterior painting offline preparation is often enough. For interior painting this can be fine, but sometimes you need to move few points by hand in booth coz the CAD usually dont show absolutely everything (think springs holding cardoors closed).
For applying the sealant, where required application precision is in similar range to manufacturing tolerances, it make sense to only prepare program structure, as you need to manually teach every point precisely on the same body that was just measured.
This would be amazing if you could do it on a car that already is completed, and could be any make/model including old cars. The trick is that the machine has to know exactly where the surface is, and that would presumably need high precision scanning.
Layers of clear coat are applied after the paint layers. To repaint / modify a car you would need to first sand away the clear coat, then apply paint over the existing paint and do color matching (paint fades over time in sunlight; clear coat reduces but doesn't stop this, so a human must blend between the freshly painted area and the existing paint), then refinish with new clear coat.
All of this is very labor intensive and requires human experience and judgement. Actually putting the new paint down is a small part of the process and there wouldn't be much to gain by automating it.
For things like putting stripes on the car, as shown in the video, vinyl decals are a much better option. A lot can be done with vinyl wraps and decals. Applying them takes some time and human expertise to evenly stretch the vinyl and apply it without bubbles/creases/warping, but the process is very straightfoward compared to painting - you just clean the car and stick on the vinyl with the help of a hair dryer / heat gun.
Even for changing the color of your entire car I'm pretty sure wrapping comes out to be cheaper than a good paint job just due to the huge amount of human labor that painting requires (also automotive paint is expensive). And there are some really beautiful colors/effects available in wraps. And a full vinyl wrap is removable - just peel it off - so you can 'undo' the 'repaint' if you use vinyl. However for decorations like stripes, if you've left the decals on for a few years then it'll look ugly if you take them off since the decals will stop the paint under them from fading but the rest of the car's paint will continue to fade from sun/UV exposure, so you'll see the decal's shape 'imprinted' into the car's paint.
Cool paint tech, they should allow you to make your own design and send it in. They could even have a ladder system that displays most popular contributions.
BMW has officially lost it. With Tesla eating the upper end of its market all they can think of is "kidney cancer" grilles and racing stripes. Folks, you need a competitive EV. Anything else you're doing is bullshit. You know it, your customers (me) know it, and your investors know it as well. There isn't currently a car in BMW's lineup that I, a BMW customer, would want, and no amount of fancy paint or juvenile racing stripes are going to change that. In fact, my next car (Cybertruck) won't have any paint at all.
I did specifically mention "competitive", which this is not. Literally, this is Blackberry against Tesla's "iPhone" - it has no chance except among the most die hard of BMW fans, and it's also goofy looking with its kidney cancer grille.
Continue? You haven't made any interesting points to start with.
The charging infrastructure is particularly odd thing to claim. The iX has the standard CCS Type 2 Combo port for Europe and it will have a CCS Type 1 Combo port for North America. It can charge at any CCS charger on any charging network.
Tesla will also finally open their chargers to all brands (as other charging networks have done for a long time):
Tesla uses CCS in Europe so those chargers are ready to go with a software update (but longer cables would be useful). Tesla will start to put CCS plugs on their North American chargers.
The non-standard charge port on the North American Teslas will be a liability going forward. Tesla will need to switch to CCS inlets at some point soon. The European Teslas are better cars because they have the standard CCS charging port and can easily charge on any CCS charger from any charging network.
It's pointless to fight the growth charging standard that the rest of the industry has switched to. The smart move is to embrace it and benefit from it.
AI and dashboard tech as well. And knowing BMW, likely pricing, too. Also BMW considers the car "done" once it's sold to you and there are hardly any updates (you can update maps for a few hundred dollars, but that's about it). For a low tech "traditional" car that's not that big a deal. For an electric though, Tesla fixes bugs and adds new functionality years after the sale. Luckily for BMW the huge and widening gap that exists between Tesla and the rest of the industry only becomes obvious once you experience one of Tesla's cars, and Tesla doesn't make that particularly easy to do due to the complete lack of the marketing effort. All the market traction they're getting is strictly word of mouth. BMWs racing stripes and half-assed EV attempts remind me of Ballmer laughing at Apple charging $600 for the first iPhone - dude simply had no capacity to see what made it worth $600, so he couldn't rip it off even if he wanted to. I bet dude felt the "brand" would solve everything, too.
I think the 'no overspray' aspect of this is probably a bit oversold, given that there doesn't appear to be any novel technology here - just close spray through a flow orifice. People who are willing to pay for this kind of service want clean, high contrast lines. Tape does a great job of this, I am skeptical that this can compare. If it does I'm sure it'll be huge though.
https://www.durr.com/en/products/paint-shop-application-tech...