5 ms·

Volkswagen’s $50B Plan to Beat Tesla Short-Circuited

15 points by wombatmobile 6y ago

wombatmobile 6y ago
The core challenge for VW, as well as for Tesla, is not just how to excel at software, as opposed to building cars, which VW is already good at. VW's ICE ecosystem works for 3rd party car component manufacturers, but VW doesn't yet know how to engineer software, or how to develop a software platform that is the basis of an ecosystem in which 3rd party devs could flourish. The challenge seems similar to the one faced by personal computers in 1980, when computing was still dominated by mainframes. At that point, the whole computer industry was an ecosystem controlled by a couple of mainframe manufacturers, and populated by a few thousand COBOL and FORTRAN programmers. The Apple ][ and CP/M computers existed, but not at a disruptive scale. Personal computers took off when the IBM PC provided a platform where 3rd party application developers could flourish, technically and commercially. The same thing happened again a quarter century later with the iPhone and Android ecosystems. What will be the enabler for EV's to grow a huge ecosystem? A monolithic Tesla? An Android or an IOS for Automobiles?
- iancmceachern 6y ago
  I recently saw a startup with just this goal, trying to remember their name...
theluketaylor 6y ago
VW stumbled pretty hard with the software for MEB, but the mechanicals are very strong. They are getting good efficiency numbers and have the toolkit to produce a ton of different EV models quite quickly. And there is no question VW is good at building cars. The real question is can VW figure out software before Tesla can figure out car manufacturing. Model Y sharing a large number of components with Model 3 is a decent start, but MEB is generations ahead on component and engineering sharing. Tesla was able to get out Model Y ahead of schedule, another good sign. Tesla's software development is pretty mature (though as an owner I'd say they are too willing to ship immature capabilities). Building a mature software engineering capability is not easy (which is why Ford has brought in so much former QNX/Blackberry talent for Mach E). Neither is building out a global manufacturing footprint that consistently delivers high volume and low defect counts.
txdv 6y ago
When Volkswagen stood in court for dieselgate, their legal language suggested that the Software devs on their own decided to write the cheat software for diesel cars. it seems that software developers do not hold very important positions in VW
sunstone 6y ago
Large manufacturer's like VW have become used to specializing in final assembly and marketing with everything else being subcontracted. With EV's a lot of things need to be done in house because the product is not yet mature. Software perhaps is the most difficult of these but clearly batteries and electric motors will also need a lot more care and attention than VW is used to.

15 points by wombatmobile 6y ago

The core challenge for VW, as well as for Tesla, is not just how to excel at software, as opposed to building cars, which VW is already good at. VW's ICE ecosystem works for 3rd party car component manufacturers, but VW doesn't yet know how to engineer software, or how to develop a software platform that is the basis of an ecosystem in which 3rd party devs could flourish. The challenge seems similar to the one faced by personal computers in 1980, when computing was still dominated by mainframes. At that point, the whole computer industry was an ecosystem controlled by a couple of mainframe manufacturers, and populated by a few thousand COBOL and FORTRAN programmers. The Apple ][ and CP/M computers existed, but not at a disruptive scale. Personal computers took off when the IBM PC provided a platform where 3rd party application developers could flourish, technically and commercially. The same thing happened again a quarter century later with the iPhone and Android ecosystems. What will be the enabler for EV's to grow a huge ecosystem? A monolithic Tesla? An Android or an IOS for Automobiles?

iancmceachern 6y ago
I recently saw a startup with just this goal, trying to remember their name...

theluketaylor 6y ago

VW stumbled pretty hard with the software for MEB, but the mechanicals are very strong. They are getting good efficiency numbers and have the toolkit to produce a ton of different EV models quite quickly. And there is no question VW is good at building cars. The real question is can VW figure out software before Tesla can figure out car manufacturing. Model Y sharing a large number of components with Model 3 is a decent start, but MEB is generations ahead on component and engineering sharing. Tesla was able to get out Model Y ahead of schedule, another good sign. Tesla's software development is pretty mature (though as an owner I'd say they are too willing to ship immature capabilities). Building a mature software engineering capability is not easy (which is why Ford has brought in so much former QNX/Blackberry talent for Mach E). Neither is building out a global manufacturing footprint that consistently delivers high volume and low defect counts.

txdv 6y ago

When Volkswagen stood in court for dieselgate, their legal language suggested that the Software devs on their own decided to write the cheat software for diesel cars. it seems that software developers do not hold very important positions in VW

sunstone 6y ago

Large manufacturer's like VW have become used to specializing in final assembly and marketing with everything else being subcontracted. With EV's a lot of things need to be done in house because the product is not yet mature. Software perhaps is the most difficult of these but clearly batteries and electric motors will also need a lot more care and attention than VW is used to.