7 ms·

Auto companies are lacking key ingredients for the future of the automobile

53 points by cocoflunchy 11y ago

alaskamiller 11y ago
The note about car manufacturers these days being system integrator is the killer point. Over 100 years now car industries have started, evolved, grown, multiplied, consolidated, failed, and experimented to today. Amazingly efficient at taking chunks of metal and plastic from around the world, turn it into a car in 17 hours in one of three or four locations in America, transport it to where you are so it can sit on a lot for another week before getting sold to you to take home. That efficiency is what makes it amazing. And that efficiency is going to be what kills it. Apple will come in, make a better ipod as they want to do and support it as they can do.
- DanielBMarkham 11y ago
  Yes. I agree. I see a lot of companies determined to outsource all of their IT work. As we all know, most of the time this fails miserably. But even sadder is when it succeeds. IT is the brain of any organization. If you succeed in distributing all the brains of your organization, all you end up being is an integrator. You've just spent a big part of your resources training technical folks in a bunch of other companies all of the important parts of your business. Success! You've turned your company into a giant brainless robot which efficiently runs on very little capital. And then you stagnate.
- w_t_payne 11y ago
  Yes. It is difficult to be innovative without the right IT infrastructure. This does NOT mean investing lots of money in IT ... but it DOES mean being careful about what tools you use and about how they integrate together, and about how they can be modified as the needs of the business grow. I.e. the organisation needs to focus on open standards and cross-platform APIs that can be built upon to meet short-term and site-specific needs rather than an over-reliance upon specific vendors or integrated and centralised toolchains that cannot be modified or built upon by individual sites and teams.
- vlehto 11y ago
  I disagree. Apple outsources too. If you need a production line, it doesn't matter too much who builds it. It takes time to build it anyhow. Both Apple and GM still produce themselves their key components. Apple makes motherboards and operating systems. GM makes engines and has chassis production in house. There is big differences, but this is not it. You could claim Apple is less agile, as they lack mass customization of hardware.
Hermel 11y ago
Software companies also lack key ingredients for the future of the automobile. The first to successfully combine the best of both worlds wins.
escherize 11y ago
Cars are kind of like tvs. My philosophy has always been to get the cheapest nav/entertainment system (or dumbest tv) and use the car speakers like a dumb pipe, because nearly every auto maker's attempt at integrating technology are completely subpar. If you expect to keep a car for more than a year or two, your best bet is aftermarket or third party components anyway. I'd say one of the few that get it is Tesla. The huge touchscreen is pretty cool, even though you lose tactile feedback, a huge nav screen with consistent ota updates is neat.
- creshal 11y ago
  > The huge touchscreen is pretty cool When I saw one last year, I laughed my ass off (and dinted my head against the too damn low ceiling – are the back seats made for dwarfs?). Whatever they're running on it, the UX is woefully outdated and looks like our one-week demos made in 2011. Not really an improvement over other car manufacturers, IMO.
- vinay427 11y ago
  I hated using a touchscreen navigation system in a previous car. The loss of tactile feedback is nothing to sneeze over, as I much prefer even a rotating dial interface which I can use while completely visually (not cognitively of course) focused on the road. This increases at least perceived safety by a large margin. An ideal combination might be a touchscreen with tactile feedback buttons, which I think has been used to limited success before. Or just voice or alternative input...
jkot 11y ago
Maybe I am wrong, but I think key innovation is to reduce cost and environmental impact. Tesla and other high tech cars are maybe cleaner to run, but have large footprint due to their complicated production. At this front leader is India and China. Tatu in India is doing stuff like gluing car parts together. And China has huge fleet of electric vehicles and recharging stations, in form of electric bikes.
- cocoflunchy 11y ago
  The total footprint is already lower than for ICE cars according to this study: http://www.ucsusa.org/clean-vehicles/electric-vehicles/life-cycle-ev-emissions http://www.ucsusa.org/clean-vehicles/electric-vehicles/life-...
  - vive-la-liberte 11y ago
    What is ICE? India, China and something else?
    - wibr 11y ago
      https://en.wikipedia.org/wiki/Internal_combustion_engine https://en.wikipedia.org/wiki/Internal_combustion_engine
    - ctdonath 11y ago
      Internal Combustion Engine.
  - Shivetya 11y ago
    does this include disposing of all these batteries? I have no doubt that some range will come strictly from batteries/capacitors for all vehicles. I do not see the current battery technology which requires nearly a thousand pounds of batteries as doing it especially for the scant range it produces. Innovation, just because it does not occur where you think it should does not mean the traditional automakers are not constantly innovation. It is the reliance on many suppliers that allows for innovation to move ahead quickly. All this work towards self driving cars came from developments made by the industry in making safer cars, from ABS to traction control to self aiming lights to airbags. Electrification, the odd thing here is that Tesla kept a 12v system when many have been trying to move all cars to 48v. I think the author is too quick to dismiss the big companies, look how fast GM turned up with the 200 mile range Bolt which should debut before Tesla's III, let alone probably reach higher production numbers Software, the author seems wholly tied up in interfaces. Why do many cars have all these dials? Assurance for the most part and a lot of tradition. Plus they aren't distracting, they are causal glance information sources whereas pretty screens and graphics likely will be distracting. Oh, as for quoting Lutz, while he may not be your cup of tea the guy knows the current industry and how it all works. He does come off like a politician at times and keeps his name in the press by being loud
rmason 11y ago
I've seen this movie before. In the seventies I had a birds eye view here in Michigan as the Big 3 totally missed the Asian threat. Now the same smug group makes jokes about Tesla (and Apple) and when it rises up to smite them they will be caught totally unaware. This time a trade embargo won't allow them to regroup.
- jacquesm 11y ago
  The car industry is a lot larger than 'the big 3' (or really, the big 2 1/2).
tiredwired 11y ago
The best selling car in America is the Ford F-series pickup truck. I can't wait to see competing trucks from Tesla and Apple.
jasode 11y ago
If one mentally redefines what a "car" will become, it can inform (and also mislead) predictions. Several industry writers have jokingly observed that a car like Tesla is really a "laptop with wheels". If we extend that reframing to say that future cars are really "robots with wheels", it doesn't look so preposterous that Tesla, Google[1], and Apple (rumoured self-driving car) could dominate Ford, GM, Toyota, etc. It seems possible that a 75% or more of a future car's R&D costs will be the sensors + computer algorithms + petabytes of cloud data integration instead of things like the tires, paint formulations, suspensions, etc. To riff on a phrase from CEO Reed Hastings, "The Goal Is to Become HBO Faster Than HBO Can Become Us" ... it seems reasonable for Tesla/Google/Apple to acquire competencies in assembling chassis with wheels faster than Ford/GM/Toyota to acquire competitive knowledge in robotics+algorithms. Time will tell. [1] robotics company acquired by Google: https://www.youtube.com/user/BostonDynamics https://www.youtube.com/user/BostonDynamics
- w_t_payne 11y ago
  The automotive industry isn't populated by idiots -- but writing safety critical software isn't the same as writing an app. It is true that there is no fundamental conflict between modern software development practices (Agile etc..) and process-maturity oriented standards like ISO15504 / ASPICE / ISO26262 / DO-178b etc... but writing the tools, build-systems etc... that let these differing concerns coexist in harmony does take some careful consideration, planning, and organisation.
  - vlehto 11y ago
    Also "chassis with wheels" is easy. Congrats, you have built a trailer. Building a car to satisfy modern consumer is not that difficult, but it isn't trivial either. And it will be expensive. Software companies are good at making software. But car manufacturers currently are some of the leading experts in computer modelling of physical world.
    - w_t_payne 11y ago
      Quite. Also, organising engineering teams on a global scale isn't trivial ... and much of the engineering process is about coordinating people, rather than specifically being concerned with the things that those people are doing.
burgreblast 11y ago
This is a great post from someone not in the industry. Car companies can't outsource production to Foxconn. Source: gesamtprojectintegrationleiter (complete project integration director) for a german car brand. After the car is "finished" setting up the production line and making the tools is a 24-36 month process. You _need_ suppliers both for sheer capacity and expertise. On my project, 17,000 parts (most with subcomponents) coming together. THEN crash, emissions, certification, etc. Easier of course to do at small scale like Tesla, but growth and capacity planning are huge topics. Tesla's run rate is what 30K cars per year? (not sure, just a guess). BMW sold 2M cars last year. Factories, suppliers, labor, paint, tooling (not the fun stuff) consumes an unbelievable number of cycles. Obviously that revolution will come from software, but there's no getting around the hardware. One of the software co's will have to buy a car company to achieve revolution and scale.
- ctdonath 11y ago
  So build a Foxconn. Someone is moving to create a billion dollar scale auto manufacturing facility in Atlanta. Seems nobody knows who, but guesses abound. Both Apple and Tesla realize this is a solvable problem and have the money to make it happen. Blog Wait But Why brilliantly detailed the the return to first principles for enormous cost savings. We've seen it before, we'll see it again. Someone did create Foxconn, after all.
  - Dwolb 11y ago
    To become a 'Foxconn' of automotive is really an interesting idea. The model of designing a vehicle might have to become more constrained which cuts right into the core of the automotive industry's design processes and car model portfolios. They're trying as hard as possible to horizontally differentiate from the competition and vertically differentiaite to access more segments. A billion dollar plant isn't really novel for the car industry - that's what it costs to bring a new facility for a model or family of models online. I think the missing link to do what you're describing (a 'flow shop' that can be very rapidly reconfigured to act like a 'job shop') would take an order of magnitude or two more investment which, given the automotive industry's single-digit profit margins, may be too risky for the incumbents. I've heard of Tesla trying to do this, but not without major difficulty (no one has ever scaled this while simultaneously building a brand new company and car and distribution network and being at the forefront of aluminum body components, etc.)
w_t_payne 11y ago
I work for one of the tier 1 automotive suppliers mentioned, and the author couldn't be more wrong about data. We have exceedingly large data sets -- equating to millions of km of driving from locations all around the globe -- and we use that data in a very disciplined and very effective manner. (My particular area of interest is in the use of off-line machine learning algorithms to analyse and optimise the algorithms that go onto the vehicle as embedded software).
- phoboslab 11y ago
  So where's the data from?
  - w_t_payne 11y ago
    We buy lots of cars, equip them with lots of different sensors, design and manufacture data recording systems to fit in them, organize configuration-management systems to manage and track changes to these test platforms, then drive them around collecting data, driven by the requirements that are placed upon our algorithms and systems, then analyse the data using both automated and manual methods. It is a time-consuming and capital intensive endeavour, no doubt about it.
dandare 11y ago
I bought new car recently and all the electronics is a bad joke. Especially the "on board computer" with "large touch enabled screen". Low resolution, dull colours, limited viewing angles, broken UI, bare minimum of functions. UX of a washing machine. Dear automakers, just open up the APIs to aircon and seatbelt sensors and leave a big hole in the middle of the front panel with USB-C connector and don't force your obsolete technology on me. Any half decent Android tablet is vastly superior to whatever "computer" you hardwired into the car back in 2007.
- yitchelle 11y ago
  Will any half decent android tablet still be functional after 10 years of vibration and exposures to vast thermal variation?
  - paulmlewis 11y ago
    Does it need to be? I think the suggestion is allow using your own tablet with a standard interface, if it breaks in a few years replace it with a newer one. That way the you're not stuck with the obsolete infotainment unit that would have been built into the car.
- danmaz74 11y ago
  But you don't entrust your life - and that of your family - to any half decent Android tablet. With cars, you can't "break things, move fast" as in some other industries.
  - marssaxman 11y ago
    I'd rather trust my life to a half-decent Android tablet than whatever insanity the carmakers are likely to have perpetrated in their firmware, but the real answer is that it shouldn't matter. Only a criminally irresponsible degree of incompetence can justify the connection of any of that fancy crap to any part of the car that has anything to do with its actual operation through any connection more substantial than power+ground.
jakobegger 11y ago
Whenever I read an article like this, with factual errors and inaccuracies in the details (eg no modern car sold today has a carburetor), I wonder how believable the big picture ideas presented in the article are? Does the author have any idea what he is talking about? Did he just make a small mistake when trying to add details to his ideas, or does the lack of subject matter knowledge reveal the writer's cluelessness?
- mywittyname 11y ago
  He clearly doesn't. Nearly every paragraph has either an inaccurate or ignorant statement that makes this really painful to read. It's easy to criticize and sound prescient when you hand-wave away reality.
vlehto 11y ago
I'm skeptical about this. It seems like the people who say "I want a self driving car" seem to be really saying they wish they didn't need a car at all. The people tho like cars often say "I want a car with less computers". What they probably really want is access to the technology they have already bought. I'm pretty sure Uber has not run it's course yet. They could almost destroy the market for self driving cars. Especially with competition. I hope that cars go to the direction of bicycles. Currently bicycles are more standardized than ever, with more options than ever. And with more competition than ever. It's hard(?) to find a new bicycle with all parts coming from same manufacturer. To me it seems this modularity is ultimately good for consumers. If capitalism works, that's where cars should gravitate too.
- padobson 11y ago
  Seems to me that Uber won't stop the market for self-driving cars, so much as validate it. They'll save a lot of money when they don't have to pay drivers anymore. Or... One person or company that owns a fleet of self-driving cars and sells them all on Uber. I think this is what the writer mean when he said self-driving cars will eliminate the need for parking lots. You'll walk out of any building in the developed world and your ride to the next place will be an driverless Uber cab.
  - vlehto 11y ago
    If someone gives me a ride with a deal that he drives and I pay the gas, how could anything compete with that? Lot's of people genuinely enjoy driving. Also bus ticket costs me 45e/month. They have schedule, which is huge asset. You could make self driving bus, but that market isn't as big. And the cost savings aren't that huge. With bus, large share of the costs come from fuel and maintenance. There are minibuses I can already order to my doorstep. https://kutsuplus.fi/tour https://kutsuplus.fi/tour And for really seldom occasions renting a car is becoming more and more easy and cheap. If it takes 15 years to develop self driving car, it's possible that 90% of the current market for it has melted away.
buildops 11y ago
This is why you have automated companies like Mobileye which provide the innovation in autos
tumba 11y ago
> ... their dashboards are (likely to be) gorgeous and > vastly improved over the mostly superfluous dials and > gauges car manufacturers think we need to see (when was > the last time you had to check your RPMs or engine > temperature?). Being aware of engine RPMs and temperature is important. I'm all for better interfaces. Personally, I drive a diesel Volkswagen and having a customizable dashboard that could provide access to additional engine parameters and sensors, such as I can access today with a laptop connected using Vag-Com (see http://www.ross-tech.com http://www.ross-tech.com) could be useful. But the key is to provide better situational awareness for the driver, not more distractions. In the US alone, more than 40k people die every year and millions are injured while operating or riding in these machines. Self-driving cars may be the future in some urban areas and on interstates, but we need to continue to optimize vehicles for drivers, not make them into more distracting spaces where people can continue to indulge their addictions to communication and media consumption.
S_A_P 11y ago
There are some good points made, and I do believe this is a solvable problem, but I also view google and apples entry into car building as naive on some levels. I am way oversimplifying but it strikes me as a plot to "silicon valley". Hooli's CEO telling the board/company all hands one day after meeting with his guru- WERE GONNA BUILD A CAR!!! Everyone looks at each other incredulously and says "umm ok, I guess we're gonna build a car". Then queue the subplot that 90% of the engineers never learned to drive and bighead is the only driver... I know this isn't the real case, and I know some brilliant people are working the problem for these companies, but I think that a better approach would be for a tech company to team up with the hardware company and leverage their combined strengths. That would probably yield a better product that is less "Jony Ive's magical mystery mobile". Shallow or not, that google car would embarrass me to ride in.
merb 11y ago
It's not the ingredients, it's the future they are missing. Car companies wanting to generate money and not the future. They lost their vision. Once upon a time a new guy will come with a vision. Hopefully he has enough money too change something. I really don't understand why we are still driving with fossil oil. That's prolly due to the lack of competetion. car manufacturers thing we will buy their cars anyway.
roflchoppa 11y ago
you know i really like my carb'd V8. Nothing like taking it out early in the morning when no-one is on the road and quite literally pouring gasoline into the motor. I feel that a cup would be more effective sometimes tho. Don't get me started on jetting configs.

53 points by cocoflunchy 11y ago

The note about car manufacturers these days being system integrator is the killer point. Over 100 years now car industries have started, evolved, grown, multiplied, consolidated, failed, and experimented to today. Amazingly efficient at taking chunks of metal and plastic from around the world, turn it into a car in 17 hours in one of three or four locations in America, transport it to where you are so it can sit on a lot for another week before getting sold to you to take home. That efficiency is what makes it amazing. And that efficiency is going to be what kills it. Apple will come in, make a better ipod as they want to do and support it as they can do.

DanielBMarkham 11y ago
Yes. I agree. I see a lot of companies determined to outsource all of their IT work. As we all know, most of the time this fails miserably. But even sadder is when it succeeds. IT is the brain of any organization. If you succeed in distributing all the brains of your organization, all you end up being is an integrator. You've just spent a big part of your resources training technical folks in a bunch of other companies all of the important parts of your business. Success! You've turned your company into a giant brainless robot which efficiently runs on very little capital. And then you stagnate.
w_t_payne 11y ago
Yes. It is difficult to be innovative without the right IT infrastructure. This does NOT mean investing lots of money in IT ... but it DOES mean being careful about what tools you use and about how they integrate together, and about how they can be modified as the needs of the business grow. I.e. the organisation needs to focus on open standards and cross-platform APIs that can be built upon to meet short-term and site-specific needs rather than an over-reliance upon specific vendors or integrated and centralised toolchains that cannot be modified or built upon by individual sites and teams.
vlehto 11y ago
I disagree. Apple outsources too. If you need a production line, it doesn't matter too much who builds it. It takes time to build it anyhow. Both Apple and GM still produce themselves their key components. Apple makes motherboards and operating systems. GM makes engines and has chassis production in house. There is big differences, but this is not it. You could claim Apple is less agile, as they lack mass customization of hardware.

Hermel 11y ago

Software companies also lack key ingredients for the future of the automobile. The first to successfully combine the best of both worlds wins.

escherize 11y ago

Cars are kind of like tvs. My philosophy has always been to get the cheapest nav/entertainment system (or dumbest tv) and use the car speakers like a dumb pipe, because nearly every auto maker's attempt at integrating technology are completely subpar. If you expect to keep a car for more than a year or two, your best bet is aftermarket or third party components anyway. I'd say one of the few that get it is Tesla. The huge touchscreen is pretty cool, even though you lose tactile feedback, a huge nav screen with consistent ota updates is neat.

creshal 11y ago
> The huge touchscreen is pretty cool When I saw one last year, I laughed my ass off (and dinted my head against the too damn low ceiling – are the back seats made for dwarfs?). Whatever they're running on it, the UX is woefully outdated and looks like our one-week demos made in 2011. Not really an improvement over other car manufacturers, IMO.
vinay427 11y ago
I hated using a touchscreen navigation system in a previous car. The loss of tactile feedback is nothing to sneeze over, as I much prefer even a rotating dial interface which I can use while completely visually (not cognitively of course) focused on the road. This increases at least perceived safety by a large margin. An ideal combination might be a touchscreen with tactile feedback buttons, which I think has been used to limited success before. Or just voice or alternative input...

jkot 11y ago

Maybe I am wrong, but I think key innovation is to reduce cost and environmental impact. Tesla and other high tech cars are maybe cleaner to run, but have large footprint due to their complicated production. At this front leader is India and China. Tatu in India is doing stuff like gluing car parts together. And China has huge fleet of electric vehicles and recharging stations, in form of electric bikes.

cocoflunchy 11y ago
The total footprint is already lower than for ICE cars according to this study: http://www.ucsusa.org/clean-vehicles/electric-vehicles/life-cycle-ev-emissions http://www.ucsusa.org/clean-vehicles/electric-vehicles/life-...
- vive-la-liberte 11y ago
  What is ICE? India, China and something else?
  - wibr 11y ago
    https://en.wikipedia.org/wiki/Internal_combustion_engine https://en.wikipedia.org/wiki/Internal_combustion_engine
  - ctdonath 11y ago
    Internal Combustion Engine.
- Shivetya 11y ago
  does this include disposing of all these batteries? I have no doubt that some range will come strictly from batteries/capacitors for all vehicles. I do not see the current battery technology which requires nearly a thousand pounds of batteries as doing it especially for the scant range it produces. Innovation, just because it does not occur where you think it should does not mean the traditional automakers are not constantly innovation. It is the reliance on many suppliers that allows for innovation to move ahead quickly. All this work towards self driving cars came from developments made by the industry in making safer cars, from ABS to traction control to self aiming lights to airbags. Electrification, the odd thing here is that Tesla kept a 12v system when many have been trying to move all cars to 48v. I think the author is too quick to dismiss the big companies, look how fast GM turned up with the 200 mile range Bolt which should debut before Tesla's III, let alone probably reach higher production numbers Software, the author seems wholly tied up in interfaces. Why do many cars have all these dials? Assurance for the most part and a lot of tradition. Plus they aren't distracting, they are causal glance information sources whereas pretty screens and graphics likely will be distracting. Oh, as for quoting Lutz, while he may not be your cup of tea the guy knows the current industry and how it all works. He does come off like a politician at times and keeps his name in the press by being loud

rmason 11y ago

I've seen this movie before. In the seventies I had a birds eye view here in Michigan as the Big 3 totally missed the Asian threat. Now the same smug group makes jokes about Tesla (and Apple) and when it rises up to smite them they will be caught totally unaware. This time a trade embargo won't allow them to regroup.

jacquesm 11y ago
The car industry is a lot larger than 'the big 3' (or really, the big 2 1/2).

tiredwired 11y ago

The best selling car in America is the Ford F-series pickup truck. I can't wait to see competing trucks from Tesla and Apple.

jasode 11y ago

If one mentally redefines what a "car" will become, it can inform (and also mislead) predictions. Several industry writers have jokingly observed that a car like Tesla is really a "laptop with wheels". If we extend that reframing to say that future cars are really "robots with wheels", it doesn't look so preposterous that Tesla, Google[1], and Apple (rumoured self-driving car) could dominate Ford, GM, Toyota, etc. It seems possible that a 75% or more of a future car's R&D costs will be the sensors + computer algorithms + petabytes of cloud data integration instead of things like the tires, paint formulations, suspensions, etc. To riff on a phrase from CEO Reed Hastings, "The Goal Is to Become HBO Faster Than HBO Can Become Us" ... it seems reasonable for Tesla/Google/Apple to acquire competencies in assembling chassis with wheels faster than Ford/GM/Toyota to acquire competitive knowledge in robotics+algorithms. Time will tell. [1] robotics company acquired by Google: https://www.youtube.com/user/BostonDynamics https://www.youtube.com/user/BostonDynamics

w_t_payne 11y ago
The automotive industry isn't populated by idiots -- but writing safety critical software isn't the same as writing an app. It is true that there is no fundamental conflict between modern software development practices (Agile etc..) and process-maturity oriented standards like ISO15504 / ASPICE / ISO26262 / DO-178b etc... but writing the tools, build-systems etc... that let these differing concerns coexist in harmony does take some careful consideration, planning, and organisation.
- vlehto 11y ago
  Also "chassis with wheels" is easy. Congrats, you have built a trailer. Building a car to satisfy modern consumer is not that difficult, but it isn't trivial either. And it will be expensive. Software companies are good at making software. But car manufacturers currently are some of the leading experts in computer modelling of physical world.
  - w_t_payne 11y ago
    Quite. Also, organising engineering teams on a global scale isn't trivial ... and much of the engineering process is about coordinating people, rather than specifically being concerned with the things that those people are doing.

burgreblast 11y ago

This is a great post from someone not in the industry. Car companies can't outsource production to Foxconn. Source: gesamtprojectintegrationleiter (complete project integration director) for a german car brand. After the car is "finished" setting up the production line and making the tools is a 24-36 month process. You _need_ suppliers both for sheer capacity and expertise. On my project, 17,000 parts (most with subcomponents) coming together. THEN crash, emissions, certification, etc. Easier of course to do at small scale like Tesla, but growth and capacity planning are huge topics. Tesla's run rate is what 30K cars per year? (not sure, just a guess). BMW sold 2M cars last year. Factories, suppliers, labor, paint, tooling (not the fun stuff) consumes an unbelievable number of cycles. Obviously that revolution will come from software, but there's no getting around the hardware. One of the software co's will have to buy a car company to achieve revolution and scale.

ctdonath 11y ago
So build a Foxconn. Someone is moving to create a billion dollar scale auto manufacturing facility in Atlanta. Seems nobody knows who, but guesses abound. Both Apple and Tesla realize this is a solvable problem and have the money to make it happen. Blog Wait But Why brilliantly detailed the the return to first principles for enormous cost savings. We've seen it before, we'll see it again. Someone did create Foxconn, after all.
- Dwolb 11y ago
  To become a 'Foxconn' of automotive is really an interesting idea. The model of designing a vehicle might have to become more constrained which cuts right into the core of the automotive industry's design processes and car model portfolios. They're trying as hard as possible to horizontally differentiate from the competition and vertically differentiaite to access more segments. A billion dollar plant isn't really novel for the car industry - that's what it costs to bring a new facility for a model or family of models online. I think the missing link to do what you're describing (a 'flow shop' that can be very rapidly reconfigured to act like a 'job shop') would take an order of magnitude or two more investment which, given the automotive industry's single-digit profit margins, may be too risky for the incumbents. I've heard of Tesla trying to do this, but not without major difficulty (no one has ever scaled this while simultaneously building a brand new company and car and distribution network and being at the forefront of aluminum body components, etc.)

w_t_payne 11y ago

I work for one of the tier 1 automotive suppliers mentioned, and the author couldn't be more wrong about data. We have exceedingly large data sets -- equating to millions of km of driving from locations all around the globe -- and we use that data in a very disciplined and very effective manner. (My particular area of interest is in the use of off-line machine learning algorithms to analyse and optimise the algorithms that go onto the vehicle as embedded software).

phoboslab 11y ago
So where's the data from?
- w_t_payne 11y ago
  We buy lots of cars, equip them with lots of different sensors, design and manufacture data recording systems to fit in them, organize configuration-management systems to manage and track changes to these test platforms, then drive them around collecting data, driven by the requirements that are placed upon our algorithms and systems, then analyse the data using both automated and manual methods. It is a time-consuming and capital intensive endeavour, no doubt about it.

dandare 11y ago

I bought new car recently and all the electronics is a bad joke. Especially the "on board computer" with "large touch enabled screen". Low resolution, dull colours, limited viewing angles, broken UI, bare minimum of functions. UX of a washing machine. Dear automakers, just open up the APIs to aircon and seatbelt sensors and leave a big hole in the middle of the front panel with USB-C connector and don't force your obsolete technology on me. Any half decent Android tablet is vastly superior to whatever "computer" you hardwired into the car back in 2007.

yitchelle 11y ago
Will any half decent android tablet still be functional after 10 years of vibration and exposures to vast thermal variation?
- paulmlewis 11y ago
  Does it need to be? I think the suggestion is allow using your own tablet with a standard interface, if it breaks in a few years replace it with a newer one. That way the you're not stuck with the obsolete infotainment unit that would have been built into the car.
danmaz74 11y ago
But you don't entrust your life - and that of your family - to any half decent Android tablet. With cars, you can't "break things, move fast" as in some other industries.
- marssaxman 11y ago
  I'd rather trust my life to a half-decent Android tablet than whatever insanity the carmakers are likely to have perpetrated in their firmware, but the real answer is that it shouldn't matter. Only a criminally irresponsible degree of incompetence can justify the connection of any of that fancy crap to any part of the car that has anything to do with its actual operation through any connection more substantial than power+ground.

jakobegger 11y ago

Whenever I read an article like this, with factual errors and inaccuracies in the details (eg no modern car sold today has a carburetor), I wonder how believable the big picture ideas presented in the article are? Does the author have any idea what he is talking about? Did he just make a small mistake when trying to add details to his ideas, or does the lack of subject matter knowledge reveal the writer's cluelessness?

mywittyname 11y ago
He clearly doesn't. Nearly every paragraph has either an inaccurate or ignorant statement that makes this really painful to read. It's easy to criticize and sound prescient when you hand-wave away reality.

vlehto 11y ago

I'm skeptical about this. It seems like the people who say "I want a self driving car" seem to be really saying they wish they didn't need a car at all. The people tho like cars often say "I want a car with less computers". What they probably really want is access to the technology they have already bought. I'm pretty sure Uber has not run it's course yet. They could almost destroy the market for self driving cars. Especially with competition. I hope that cars go to the direction of bicycles. Currently bicycles are more standardized than ever, with more options than ever. And with more competition than ever. It's hard(?) to find a new bicycle with all parts coming from same manufacturer. To me it seems this modularity is ultimately good for consumers. If capitalism works, that's where cars should gravitate too.

padobson 11y ago
Seems to me that Uber won't stop the market for self-driving cars, so much as validate it. They'll save a lot of money when they don't have to pay drivers anymore. Or... One person or company that owns a fleet of self-driving cars and sells them all on Uber. I think this is what the writer mean when he said self-driving cars will eliminate the need for parking lots. You'll walk out of any building in the developed world and your ride to the next place will be an driverless Uber cab.
- vlehto 11y ago
  If someone gives me a ride with a deal that he drives and I pay the gas, how could anything compete with that? Lot's of people genuinely enjoy driving. Also bus ticket costs me 45e/month. They have schedule, which is huge asset. You could make self driving bus, but that market isn't as big. And the cost savings aren't that huge. With bus, large share of the costs come from fuel and maintenance. There are minibuses I can already order to my doorstep. https://kutsuplus.fi/tour https://kutsuplus.fi/tour And for really seldom occasions renting a car is becoming more and more easy and cheap. If it takes 15 years to develop self driving car, it's possible that 90% of the current market for it has melted away.

buildops 11y ago

This is why you have automated companies like Mobileye which provide the innovation in autos

tumba 11y ago

> ... their dashboards are (likely to be) gorgeous and > vastly improved over the mostly superfluous dials and > gauges car manufacturers think we need to see (when was > the last time you had to check your RPMs or engine > temperature?). Being aware of engine RPMs and temperature is important. I'm all for better interfaces. Personally, I drive a diesel Volkswagen and having a customizable dashboard that could provide access to additional engine parameters and sensors, such as I can access today with a laptop connected using Vag-Com (see http://www.ross-tech.com http://www.ross-tech.com) could be useful. But the key is to provide better situational awareness for the driver, not more distractions. In the US alone, more than 40k people die every year and millions are injured while operating or riding in these machines. Self-driving cars may be the future in some urban areas and on interstates, but we need to continue to optimize vehicles for drivers, not make them into more distracting spaces where people can continue to indulge their addictions to communication and media consumption.

S_A_P 11y ago

There are some good points made, and I do believe this is a solvable problem, but I also view google and apples entry into car building as naive on some levels. I am way oversimplifying but it strikes me as a plot to "silicon valley". Hooli's CEO telling the board/company all hands one day after meeting with his guru- WERE GONNA BUILD A CAR!!! Everyone looks at each other incredulously and says "umm ok, I guess we're gonna build a car". Then queue the subplot that 90% of the engineers never learned to drive and bighead is the only driver... I know this isn't the real case, and I know some brilliant people are working the problem for these companies, but I think that a better approach would be for a tech company to team up with the hardware company and leverage their combined strengths. That would probably yield a better product that is less "Jony Ive's magical mystery mobile". Shallow or not, that google car would embarrass me to ride in.

merb 11y ago

It's not the ingredients, it's the future they are missing. Car companies wanting to generate money and not the future. They lost their vision. Once upon a time a new guy will come with a vision. Hopefully he has enough money too change something. I really don't understand why we are still driving with fossil oil. That's prolly due to the lack of competetion. car manufacturers thing we will buy their cars anyway.

roflchoppa 11y ago

you know i really like my carb'd V8. Nothing like taking it out early in the morning when no-one is on the road and quite literally pouring gasoline into the motor. I feel that a cup would be more effective sometimes tho. Don't get me started on jetting configs.