Problem with Modern Cars Becoming Collectible
08-25-2017, 10:06 AM
#1
Problem with Modern Cars Becoming Collectible
So, I have this ACR (08)...some might disagree, but it's fairly modern. It has a complicated engine (cam in cam) and an ecu that was just recently cracked. Right now, it's having electrical issues and it got me thinking (only tangentially related to the issues i'm having with my car)...how are cars going to fare in 20-30-50 yrs from now with all these 'modern' (by then super old) electronics?
It seems, the most important part to be able to get ahold of is an ECU. I know that oems keep this stuff around for a while (no idea how long...and i doubt it's standard), but at some point either it will get lost in the dust bin of time or the oem might actually decide..i'm not interested in keeping all this old code and hardware around. While, I have been thinking about it in terms for collectible cars...it may be even more important for your avg car to continue to run in developing countries, etc.
So...my idea would be to gather FPGA (field programmable gate arrays) code for the various cpus that are used (ARMs, TI, Freescale, Renesas, etc) and get the code and updates for each model/trim and put it in a repo and either open source or charge for it. I know that oems might have an issue (they like to encrypt them), but after some amount of time they could allow it be unencrypted and hosted either publicly or behind a paywall....think when classified docs become declassified after some amount of time.
I personally have a lot of connections in the automotive world, but I need to spend some time to understand how oems deal with this now, their plans for the future (assuming they have any..you might be surprised), and if this would be something that would even interest them as more than likely I would need their help and cooperation.
Anyway, just wondering what people think. If anyone knows a lot about this subject I would love to hear from you. Thanks.
It seems, the most important part to be able to get ahold of is an ECU. I know that oems keep this stuff around for a while (no idea how long...and i doubt it's standard), but at some point either it will get lost in the dust bin of time or the oem might actually decide..i'm not interested in keeping all this old code and hardware around. While, I have been thinking about it in terms for collectible cars...it may be even more important for your avg car to continue to run in developing countries, etc.
So...my idea would be to gather FPGA (field programmable gate arrays) code for the various cpus that are used (ARMs, TI, Freescale, Renesas, etc) and get the code and updates for each model/trim and put it in a repo and either open source or charge for it. I know that oems might have an issue (they like to encrypt them), but after some amount of time they could allow it be unencrypted and hosted either publicly or behind a paywall....think when classified docs become declassified after some amount of time.
I personally have a lot of connections in the automotive world, but I need to spend some time to understand how oems deal with this now, their plans for the future (assuming they have any..you might be surprised), and if this would be something that would even interest them as more than likely I would need their help and cooperation.
Anyway, just wondering what people think. If anyone knows a lot about this subject I would love to hear from you. Thanks.
08-25-2017, 10:34 AM
#2
It is definitely a concern, although I think the rate that ECUs go bad will be less than the rate at which cars are junked which can donate good ECUs.
Great example of this: All those McLaren F1s worth $10M+ currently rely on an ancient Compaq laptop. They're working on a modern interface, but at this time it isn't complete and they're scouring the world for this specific type of old laptop:
deadspin-quote-carrot-aligned-w-bgr-2<\/title><path d="M10,3.5l3-3,3,3Z" style="fill:%23fff;stroke:%23fff"/><path d="M0,3.5H10l3-3,3,3H26" style="fill:none;stroke:%231b3a4d"/><\/svg>')}.f_branding_on.blog-group-deadspin .editor-inner.post-content .pu
Great example of this: All those McLaren F1s worth $10M+ currently rely on an ancient Compaq laptop. They're working on a modern interface, but at this time it isn't complete and they're scouring the world for this specific type of old laptop:
deadspin-quote-carrot-aligned-w-bgr-2<\/title><path d="M10,3.5l3-3,3,3Z" style="fill:%23fff;stroke:%23fff"/><path d="M0,3.5H10l3-3,3,3H26" style="fill:none;stroke:%231b3a4d"/><\/svg>')}.f_branding_on.blog-group-deadspin .editor-inner.post-content .pu
08-25-2017, 11:17 AM
#3
How long are auto manufacturers on the hook for parts once a car is no longer being manufactured?
Sounds like a big project, but could have a big pay day in future.
Sounds like a big project, but could have a big pay day in future.
08-25-2017, 11:44 AM
#4
10 years, as I understand it, is as long as they are required to maintain and supply parts.
08-25-2017, 11:53 AM
#5
If you had to pick a few cars that was made between 2007 and 2017; which models (beyond the 200k+ new cars) do you see as the most collectible in 2037? Where is the smart money going?
08-25-2017, 03:30 PM
#6
This will be a challenge, and will likely need to target high demand, high dollar cars that predictably will need the parts you're trying to sell. I remember a local machining company here making some kind of bracket or trim piece or something for Rolls Royces that apparently get damaged/failed all the time, and when they started getting orders for the same thing they realized there was a market for them. So the dude made batches and made decent money for it.
Going back to this, in the case of McClaren, the demand for those computers isn't that desperate, fact is there isn't that many of those cars that require that part and they do have the laptops. They recognize that they need to update it but nonetheless, had these cars been say a Toyota Corolla and they got mandated to fix a problem like that, rest assure the problem will be fixed practically overnight. I liken this back to Toyota Priuses, when the restraint systems on those cars triggered an airbag light or when the hybrid system was interrupted for repairs, only the dealer had the special tool to reset these lights because of the proprietary dongle you needed mostly. Took less than a year before black market kits were available (dealers were charging 150 a pop to plug in a machine and essentially hit a button or two, you think independent shops may want a piece of that) and soon after legit devices.
Additionally, there are numerous independent engineer type firms that can be commissioned to make specific connectors, circuits, etc, and actually modern ECUs and such would make it easier since alot of things are becoming more standardized or familiar. So would it be worth stocking physical ECUs etc, probably not, nor is the code since you can rip it out of another ECU. I do believe in the future, just like how we have transmission/engine rebuilding specialists, we will have electric motors and circuit rebuilders, the type of people like those vendors you see on Wheeler Dealers rebuilding digital gauge clusters.
Side note, often times ECU fails because the built in battery goes bad or there was a short that if push comes to shove can be repaired, and those companies already exist. So unless these things were made of plantinum and are getting stolen for their physical value, or if that particular car keeps crashing and burning to the ground, there should be enough ECUs to cover the demand.
Going back to this, in the case of McClaren, the demand for those computers isn't that desperate, fact is there isn't that many of those cars that require that part and they do have the laptops. They recognize that they need to update it but nonetheless, had these cars been say a Toyota Corolla and they got mandated to fix a problem like that, rest assure the problem will be fixed practically overnight. I liken this back to Toyota Priuses, when the restraint systems on those cars triggered an airbag light or when the hybrid system was interrupted for repairs, only the dealer had the special tool to reset these lights because of the proprietary dongle you needed mostly. Took less than a year before black market kits were available (dealers were charging 150 a pop to plug in a machine and essentially hit a button or two, you think independent shops may want a piece of that) and soon after legit devices.
Additionally, there are numerous independent engineer type firms that can be commissioned to make specific connectors, circuits, etc, and actually modern ECUs and such would make it easier since alot of things are becoming more standardized or familiar. So would it be worth stocking physical ECUs etc, probably not, nor is the code since you can rip it out of another ECU. I do believe in the future, just like how we have transmission/engine rebuilding specialists, we will have electric motors and circuit rebuilders, the type of people like those vendors you see on Wheeler Dealers rebuilding digital gauge clusters.
Side note, often times ECU fails because the built in battery goes bad or there was a short that if push comes to shove can be repaired, and those companies already exist. So unless these things were made of plantinum and are getting stolen for their physical value, or if that particular car keeps crashing and burning to the ground, there should be enough ECUs to cover the demand.
08-25-2017, 05:03 PM
#7
darcy
Trending Topics
08-27-2017, 08:52 AM
#8
low production number cars with interesting options, and lower miles (<50K)- eg. mustang 350/R, 1LE SS camaro, of course, the s2000, STi RA, USDM 1st gen TypeR, Trackhawk Jeep, Hellcat, Fancy porsche's (Boxster spyder, GT3). And remember this, where available, WITH A MANUAL TRANSMISSION- always choose 3 pedals if you hope to have some skin at the end.
darcy
darcy
08-27-2017, 12:23 PM
#9
I'm having a hard time understanding what market you're trying to capture? Sounds like you want to furnish FPGAs for potentially failing ECUs in the long-term future. A few questions that crosses my mind:
1. Who would troubleshoot the ECU to isolate the root cause to the FPGA?
2. FPGAs are less likely to fail than say the other ancillary EEE components. Unless you can provide a service to repair a failing ECU to a functioning one, I see no value here.
3. In addition to the VHDL (or whatever it may be) you'll need the compiler and all supporting software items
4. Having the code by itself without the hardware is useless. Are you planning on bridge-buying FPGAs as they become obsolete?
From what you described, I don't think this is a very good business case.
1. Who would troubleshoot the ECU to isolate the root cause to the FPGA?
2. FPGAs are less likely to fail than say the other ancillary EEE components. Unless you can provide a service to repair a failing ECU to a functioning one, I see no value here.
3. In addition to the VHDL (or whatever it may be) you'll need the compiler and all supporting software items
4. Having the code by itself without the hardware is useless. Are you planning on bridge-buying FPGAs as they become obsolete?
From what you described, I don't think this is a very good business case.
08-27-2017, 12:30 PM
#10
I'm having a hard time understanding what market you're trying to capture? Sounds like you want to furnish FPGAs for potentially failing ECUs in the long-term future. A few questions that crosses my mind:
1. Who would troubleshoot the ECU to isolate the root cause to the FPGA?
2. FPGAs are less likely to fail than say the other ancillary EEE components. Unless you can provide a service to repair a failing ECU to a functioning one, I see no value here.
3. In addition to the VHDL (or whatever it may be) you'll need the compiler and all supporting software items
4. Having the code by itself without the hardware is useless. Are you planning on bridge-buying FPGAs as they become obsolete?
From what you described, I don't think this is a very good business case.
1. Who would troubleshoot the ECU to isolate the root cause to the FPGA?
2. FPGAs are less likely to fail than say the other ancillary EEE components. Unless you can provide a service to repair a failing ECU to a functioning one, I see no value here.
3. In addition to the VHDL (or whatever it may be) you'll need the compiler and all supporting software items
4. Having the code by itself without the hardware is useless. Are you planning on bridge-buying FPGAs as they become obsolete?
From what you described, I don't think this is a very good business case.
I think it would be best as open cource project. Also, how do you expect to see cars built 10 years ago, run in 30-40 years? I'm making an assumption that is something people will want. I can't see oems stock piling old hardware, compilers, code and supporting it forever.
Edit: one more thing, you don't need the old cpus, fpgas, etc. you need custom harnesses that fit the pattern of the oem equipment. Even the original fpga layout coujd be stored. So a modern fpga could effectively be turned into an old fpga, to run the old uProcessor code, to run the oem code and you put it all in a custom chassis that fits into the oem ecu mount.