Visit the forum instructions to learn how to post to the forum, enable email notifications, subscribe to a category to receive emails when there are new discussions (like a mailing list), bookmark discussions and to see other tips to get the most out of our forum!
OSE Car Lifetime Cost Comparison
  • I made Lifetime Costs Comparison of the OSE Car:

    To buy the OS Car is about 3x cheaper, do DIY is about 5x cheaper.

    These calculation will be published in the upcoming OSE Video and would be happy to receive some feedback on the numbers.

    Thank you,
  • 15 Comments sorted by
  • One huge factor missing from the calculations is that the systemic corollary of open-source manufacturing is the elimination of vehicle-dependence and the consequent vast reduction in overall production volumes and distance travelled. The current transportation scenario is based on interests so hostile to the open-source/diy approach that it is senseless to think in terms of a direct substitution of hardware. If one thinks in terms of an alternative scenario many of the assumptions must change. Arguably they become very difficult to estimate.

    For instance, the comparison is not so much between 50mpg and 109mpg for the same annual mileage and the same cost per gallon. I should propose that a 90% reduction in annual mileage might be perfectly feasable. The price, however, that a local farmer might charge for E96 run up in batches in his 3" column still over the year is harder to anticipate. And then I submit that the car might more likely end up being built in 30mpg tune than 109mpg tune. A scenario that has very little need for cars would not necessarily be without the desire for cars - though I maintain that it would be small - and one must ask if one would bother having a car at all if it was going to be a puritanically austere one. In fact I see little place for the OSE/Wikispeed sort of brief in that scenario: imagining myself in the situation I'd rather opt for a passenger-bodied version of a farm truck.

    Likewise I should conceive the costs at various points in time slightly differently: In the first year I might have spent a bit more, depending on what I might incorporate on my own peculiar take on the design. I should summarily ignore the crash module at Year 10, and would be more inclined to use the word "aspiration" than "innovation" for the new, or radically rebuilt, engine. This is because I should like to see non-progressive innovation that cumulatively widens our technological palette in place of our current, linear, progressive "future machine", a creature of patent law.
  • Hi Ned,

    Thank you for the comment. Yes, in this case I am considering ONLY the price from the view of the customer! I am not considering the production cycle which would radically cut much more costs.

    On mpg - I have no choice but to stick with the official data - 109 mpg. Cost over the year will change and this is totally clear, but we need a MODEL to VISUALIZE the lifetime costs. For sure after 10 years this model may be different, but it is hard to create adaptive model for the future, that's why I am doing it in the most simple way. The most important factor in this model is the COMPARISON between events happening with two cars, rather than ACCURACY of what could happen. The focus is on the car in this case, a farm truck cost comparison is possible, but we don't have the data and it will not go into the planned video. Years are also chosen as an example. Crash Module is just one of the meny examples of innovation that could happen with corresponding COST COMPARISON. The concrete innovation itself is not important, there could be many other type of innovations, but we HAVE the data for this concrete example and it can serve us as a model. With these thoughts in mind, would like to suggest some improvements?
  • Apart from pointing out the hypothetical nature of the exercise I think it's quite useful. In a sense it's a worst-case scenario.
  • Ummm...not to be "that guy" but it looks like your "cost comparison" assumes the person with the industrial car will buy two brand new cars in the same time they have the SGT01. That seems a bit skewed, particularly when you include DIY cost reduction options for the SGT01 car. If the person is a DIY kind of guy, then he could probably find it within himself to keep his car running for 10 years, or at least buy a used car instead of a new car. Why would you assume someone buys a new car every 10 years in the 100 year comparison but assume they buy a new car every 5 years in the 10 year comparison?

    I'm just gonna get this off my chest now: Wikispeed's SGT01 is not an appropriate substitute for the GVCS passenger car. Yes, they designed it to be modular, yes it is a good example of appropriate technology, yes they agreed to open source it, but it was still designed without any of the same constraints the GVCS has. Personally, I think the Wikispeed/OSE collaboration should remain just a collaboration.

    I sympathize with the need to figure out some sort of cost comparison. It is far from straight-forward. I suppose, given all the unknowns, one person's guess is as good as another's. 
  • this is an older message which somehow wasn't sent few weeks ago:

    Hi Matt,

    for 10 years the goal is at the end of these 10 years to have a working car. That's why 2 industrial cars are calculated in this cost comparison. Same with the 100 years, the goal is after that again to have a working car. Is it now more clear?

    Yes, you have right that I haven't considered the case Industrial+DIY. I don't plan to include it, because currently it is also harder to DIY because of the lack of knowledge and technology designed not to be easily repaired.

    Joe Justice confirmed that the data is exactly as Wikispeed expects.
  • I updated the calculations, I forgot few considerations like acquiring a new OS Car after its Lifetime and created two cases for a Lifetime of 100 and 33 years:

    Comments are welcome. The data will be presented in 2 weeks internationally.
  • What about it is data? The Wikispeed car is just a prototype, so any numbers describing its performance over any span of years are hypothetical, right?

    I don't see how this is any more useful for evaluating the idea than just saying, "It's designed to be more efficient. So, you know, we assume it will be." Is there really a need to assign numbers to the speculation? 

    Of course, I'm not saying to avoid using numbers if someone higher up said to "get" them. I suppose numbers at least show that we can do math. I just personally distrust any big complicated formulas that are fill in with guesswork. 

    Interestingly, I was just putting together a blog post on how open source hardware companies can turn a profit. It turns out they're subject to the same rules any other business is. The "2 to 4 times cheaper" result in the spreadsheet looks like it's at risk of being eaten up by the profit margins necessary for one or more companies to stay in business. No one ever budgets what something actually turns out to cost. It always increases 50-100% when the reality of maintaining adequate stock, and satisfying customers, and paying employee healthcare, and allowing retailers to turn a profit, etc all pile up. Granted, part of the idea is to come up with a NEW way of doing business...I guess we'll see what happens.

    Also, it looks like y'all low-balled some of those numbers. Why are the "labor" costs $0? Someone has to do the work and their time is going to cost something. The odds of the person doing the work making LESS than an auto mechanic are pretty low (they already don't make much), so it seems like the labor costs should be similar. Marcin has been tossing around estimates like $5,000 of economic value per person per day. Assuming an 8 hr day that means each person's time is worth $625 (more if they get it done in less than 8 hours). When you add up the "savings" of building your own car for 100 years don't you need to add up the costs of not doing other things? Not to mention that the labor costs of a mechanic are burdened (overhead is added in) but if you're doing the work yourself you most likely aren't maintaining a full-time auto shop, so every time YOU do it you have to relearn some stuff and maybe get some new tools, which increases the amount of time you're spending working on your car. Sure, a washing machine cost money to buy and replace, but it frees you up to do much more valuable work than wash clothes, so you can earn the money to pay for the washing machine and other things. 

    It would make sense to argue that the open source vehicle will require fewer tools and less knowledge to maintain/upgrade, so THAT will save labor costs over time since you can pay someone even less money to do it. But, again, the amount saved is speculation.'s an interesting thing to try and apply numbers to, but my suggestion is still to just admit that we're guessing and move on. Well, I dunno, I suppose some people will just take the projected savings at face value. Maybe there's something to taking advantage of that.
  • Hi Matt, yes, we don't have the final numbers and I am offering
    an example of what it could be. I want to present this example because I
    believe there is some truth in there. The more you help me to make it
    more realistic, the better :)

    We have a very good discussion about that in OSE Germany:!topic/ose-germany/L4eXxWyg5D8
    Because of that I am considering now the case for 20 years, instead of 100.

    Your main points:
    1) "It always increases 50-100% when the reality of maintaining adequate
    stock, and satisfying customers, and paying employee healthcare, and
    allowing retailers to turn a profit, etc all pile up
    - This is
    not considered in the calculations. Only materials + labor costs. I
    don't want to include these because it gets too complicated. As you say
    we need new WAYS to do business so that we can lower also these costs.

    2) "y'all low-balled some of those numbers. Why are the "labor" costs $0?"
    - There were 2 actions with $0.
    One was about planned obsolescence -
    it is possible that for a product for which there is no planned
    obsolescence the user does not have any additional costs. But to avoid
    questions and doubts like this, I added $500 maintenance costs and
    lowered the planned obsolescence costs to $3000 per 10 years.
    The other was exchanging the engine
    - well it is so simple, that it takes less than 5 minutes work. Again
    to avoid questions like this I added a symbolic price of $20 for this

    I would say, let's forget about the 100 years example and consider the 20 years. Where are its weak points?
  • "it is possible that for a product for which there is no planned obsolescence the user does not have any additional costs"
    >>> like "no additional costs" always activate my skeptic gland. The way I see it, a more-or-less DIY vehicle is going to be loading most of the cost/time onto the sustainment part of the lifecycle. When I buy a vehicle from Toyota I expect it to last 10-20 years with minimal maintenance because Toyota has sunk a billion dollars of world-class development and testing into the vehicle before they delivered it. We are doing the exact opposite. We're developing and testing a vehicle with the absolute minimum expertise, time and budget and we're also designing the thing to ultimately be produced at the lowest possible cost. That means all the time and money we're not spending improving the design and the production process is going to be paid out over the life of the vehicle as it breaks or needs upgrading (expected and unexpected). Picture Microsoft Windows with its steady stream of updates/patches, only in hardware. I don't believe that a vehicle built by amateurs, out of stock materials, using commodity tools, is going to survive regular use and not incure "any additional costs." That just seems backwards. I don't know how much it will end up costing, but it will HAVE to cost more than a COTS vehicle. Unless you guys are assuming that maintenance/upgrade on COTS vehicles is so dramatically overpriced that it will always cost more than maintenance/upgrade on a kit car, even when the latter requires more maintenance/upgrade than the former.

    "exchanging the engine - well it is so simple, that it takes less than 5 minutes work. Again to avoid questions like this I added a symbolic price of $20 for this service."
    >>> MAYBE it will eventually take 5 minutes. At the moment swapping out the engine in the Wikispeed car requires removing the giant body structure, at least one interior module, at least one crush zone, and then you need tools/space/expertise to deal with the axles and pallet jack and whatnot. And that's only the stuff I know about without ever having seen the engine module in person. I'm sure there's more. Of course, you have to do all that in reverse with the new engine. However, that's not the point. The point is that someone had to build the engine module. THAT's the part that needs to be priced out. Either the engine needs to be built into the module, or the module needs to be built around the engine. Someone also has to transport the thing. Leaving aside how big and heavy it is, you can't ship things with fluids in them, so swapping engines also requires you to drain all the fluids (oil, coolant, etc) and then put all of it back in. How many people are willing to own a pallet jack just to swap out the engine in their car sometimes? I'm not even sure if you can get something that heavy delivered to your house without dealing personally with the delivery service and being present to accept delivery. Same with getting rid of it. 

    My point is that the estimate seems to be leaving out a lot of complications that are going to cost time and money. Sure, many people who see the estimate won't know enough about the context to question it, even if they do have the inclination. But if you want the estimate to survive scrutiny from people who do know something about the inevitable complications you'll have to at least not put $0 there. Or maybe do a much better job of explaining why certain costs were left out.

    Ultimately it seems like maybe the problem is that you're comparing apples to oranges. It will be a LONG time (a decade absolute minimum, probably more like two or three minimum) before owning an open source, paradigm-breaking vehicle will be anything close to normal. The turnover in our nation's vehicle fleet happens on a scale measured in decades, and that's not including the time needed to develop the technology and integrate it into the existing regulatory structure. So, for estimates to mean anything TODAY, you need to compare the Wikispeed car to its closest counterpart: kit cars. 

    That's really what it is. It's a kit car. It's a unique vehicle, designed from the ground up to be built and/or maintained by its owner, who has to have the disposable cash, free time, and interest/inclination to have a full-sized toy car around. I'm involved in enough open source hardware projects to be totally convinced that the first thousand people who own Wikispeed's cars are going to be volunteering to be part of their R&D team. I would do it, but I don't have enough money to maintain two cars (one reliable one to get me to work and one developmental one to break-fix-break-fix-break-etc).

    So, you shouldn't be comparing Wikispeed's car to a regular commercial car. You should be comparing it to a regular KIT car. Technically, to a BRAND NEW kit car design that has never been used in real life before. I'm sure the kit car industry, or at least some old guy in it, would be a good source of information on how much they actually cost. These might be some good places to start:
  • I think it's hard to really calculate the value of a car which hasn't been proven over a few hundred thousand miles. Doing a hard comparison could potentially hurt the credibility of all the other GVCS machines because there is very large market for cars and even more people who are knowledgeable about cars.

    Personally, buying cars off of craigslist is the most economical way to go. The problem with them however is they have problems. Brakes, cooling, interior, fuel injectors, etc, etc. Unless you have the time and knowledge to fix these things, they get expensive fast. The real cost to fix these things is (1) paying a mechanic, and (2) time. Down-time, taking it to get fixed, breaking down. The cost of the parts isn't actually that bad usually. It's the constant hassle. [the other problems with craigslist cars is fuel efficiency]

    An OS car would presumably be easy to diagnose and fix. I think that's the primary value. After 10 years of both closed and open source vehicles working ok, the vehicle starts having problems. With a closed source car this becomes a major pain. With an open source car, just the amount of documentation, support, and modularity make these repairs cost $200 - $300 and a couple hours, as opposed to $400 - $700 and several days.
  • >Comments are welcome.
    Okay, I'll step up to the plate. As usual, I'll be blunt.
    > The data will be presented in 2 weeks internationally.
    I think describing one's hopes as 'data' hurts credibility in the long run, though in the short run it does generate enthusiasm. If the object is to get folks to click the Donate Now button, this cost comparison may be effective. But just between us chickens, it might be helpful to do a similar analysis, on a different page perhaps, sans the spin.
    I would say, let's forget about the 100 years example and consider the 20 years. Where are its weak points?
    If it's meant to convince, it's strong as is. If it's meant to inform, then weak points abound. The comparison of an engine swap: $2500 for the CM, zero dollars for the OS/Wikispeed. The assumption is that the OS/Wikispeed person is smart and skilled enough to build their own OS car, and the CM owner isn't smart/skilled enough to even install an engine. Also, there's no comparison of cost of materials--no mention of materials at all--when the OS/Wikispeed engine module will be considerably more expensive than a CM engine. For the CM car, an engine includes an engine. For the OS/Wikispeed car, the zero labor "OSE/Wikispeed - 4 bolts, slide the engine module out" engine includes an engine, a transaxle, an electrical system, a fuel system, and plenty more. 
    Ditto the service cost comparison, which indicates the OS/Wikispeed builder is smart enough to do all service and upgrades for nothing, and the CM owner is too dumb to even change the oil.
    Quoting from the spreadsheet:
    we assume 3000 costs inflicted by planned obsolescence on CM
    Well, planned obsolescence may mean different things to different people, but if you mean it the way Vance Packard meant it, I think it was real among the Big Three back in the '50s, but the auto industry is worldwide and too competitive nowadays for...I mean come on now, the idea that modern car companies can devalue their own cars 12% via planned obsolescence without losing market share to those that don't, it's kind of stretching the conspiracy theory, isn't it?
    I would say, let's forget about the 100 years example...
    You can say it, but as long as that example is on the wiki and "The data will be presented in 2 weeks internationally" I think it's fair to note the 100 years example claims a total of $75,000 of planned obsolescence costs against the CM car, in the Action item of "Planned Obsolescence (any example)".
    > The CM's 3rd and 13th year $5000 "Broke (add new body)" fee is not justified. CM cars do not average $500 worth of body damage a year.
    > "In 5 years, the average American will trade in and buy their next car, if they did the same wasteful behaviour..."
    If the average American sells their used car every five years, then the average American buys a used car every five years. If it's wasteful behavior for the seller, it's frugal behavior for the buyer. The OS Car Lifetime Cost Comparison doesn't compare the cars as much as it compares the owners. OS/Wikispeed owners are smart and frugal, CM owners are dumb and wasteful. OS/Wikispeed owners keep their cars for 20 years and sell them for half of what it cost to build them, CM owners get rid of their cars in 10 years (according to the chart) and also sell them for half of what they cost when they were new. If this were a lifetime comparison of the cars, then years 11-20 for the CM would be as operated by the second owner. Who apparently is smarter and more frugal than the first owner; smart and frugal enough to drive a 10 year old car...though not nearly as smart and frugal as the OS/Wikispeed owner in the 100 years comparison; that OS/Wikispeed owner drives his car for a hundred years, spends less than $100 a year on maintenance and upgrades, and gets half his initial investment back when he sells it. The dumb CM owner sells early and loses more than $100 a month on depreciation, while the smart OS/Wikispeed owner keeps it in the family for generations (his grandkids will see the odometer hit its first million miles in 2087), while it depreciates at $6 a month.
    not to mention cost of fuel, at 50mpg for Toyota vs. 109mpg combined for OSE Car
    Yeah, we probably shouldn't mention that. It may be a bit early to call that 'data'.
     I want to present this example because I believe there is some truth in there.
    Nikolay, I too believe there is some truth in there. I think the recent spreadsheets are more truthful than the Wiki, but I don't think it's my place to update it. I also think if the non-truth were removed, the presentation would be more effective, and I further think that if the assumptions were removed you'd have a good solid core of unarguable truth.
  • guys, this statistics is creating too much troubles! Hahaha :)

    The fact is that I personally cannot create it in the most transparent and knowledgeable way. We need actual transparent data and more systematic analysis to create such statistics. As many of you say, this may take at least few more years.

    For the sake of quality and not willing to engage in such discussions, I deleted the Wiki page and will not present this data in future events. Nevertheless I will communicate the advantages of the OS Car as they are right now, without the concrete data and comparing it with other cars.
  • Nevertheless I will communicate the advantages of the OS Car as they are right now...

    That sounds encouraging. What -are- the advantages of the OS Car right now?

    > ...without the concrete data and comparing it with other cars.

    I think concrete data will enhance your communication of the advantages. "Concrete data" is what makes your communication informative, lack thereof makes it an opinion piece. I value your opinion and read your posts (and wiki contributions) with interest. The problem with comparing it (the OSE/Wikispeed car) with other cars is in presenting the comparison as fact, instead of identifying what are facts and what are opinions...and if you want to get deeper, identifying which opinions are based on experience, which are faith-based, and which are wishful thinking. All those opinions have their place, and all can be useful to the reader, but only if a reader can tell which is which.

    I'm looking forward to the video, and to learning the advantages of the OS Car as they are right now.
  • I have concrete data from Joe, but it is not yet transparent, what
    source? links enough. Also what is missing is the systematic analysis
    what data should go into the comparison.

    The advantages of the Wikispeed car are communicated on 4 pages under "The Car":
  • To be fair, the last time I saw it discussed on the email list Joe said that the car didn't actually get 100mpg or 5-star crash safety in actual testing. Granted, they were testing prototypes, not the finished car, but all that "data" (except maybe weight and top speed) is projected. 

    Wikispeed's car is a great idea, but it's still in the womb. Maybe the playpen. The point is that it's still developmental. It hasn't even been demonstrated in a relevant environment. I'd say the Wikispeed car is at TRL 3 (maybe 4), but definitely not 5. 

    It's important to talk about the project in the context of related projects. It's not like this is the first time anyone has ever invented a new car, or a new piece of hardware. I suggest simply stating what we've got and not reaching too far. Speculate away, but clearly label it as speculation. Something as simple as "we estimate..." or "if the car performs as expected..." should be enough. 

Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

Login with Facebook Sign In with Google Sign In with OpenID Sign In with Twitter

In this Discussion