Book Excerpt: Tesla’s vertical integration is a radical change in strategy
One of the many ways in which Tesla does things differently than other automakers is its tendency towards vertical integration. The industry giants have become aggregators, sourcing components from vast chains of suppliers all over the world. However, as Greg Reichow, Tesla’s former VP of Production, explained in a recent article, the Californian upstart builds many of its own battery, power electronics, and drivetrain systems, to say nothing of more mundane components - in some cases, it has even made its own seats. Some have estimated that, of the over 5,000 components in a Tesla vehicle, around 80% are made in-house.
Above: The Model S being built at Tesla's factory in Fremont, CA (Image: Inhabitat)
This wasn’t always the philosophy at Tesla - on the contrary, when the Roadster was designed, the young company followed precisely the opposite strategy, as Tesla co-founder Marc Tarpenning explained to me when I interviewed him for my book, Tesla: How Elon Musk and Company Made Electric Cars Cool, and Remade the Automotive and Energy Industries. The following is adapted from the book, which has been totally revised and updated, and is now available as edition 3.0.
To build the Roadster, the Tesla team began with powertrain technology from a company called AC Propulsion, and for many of the other bits, they turned to a sports car maker that they all admired - Lotus. Tesla ended up licensing various technologies from Lotus, and contracted with the elder company to manufacture the cars in its legendary factory, with Tesla supplying most of the parts.
“To get a car onto the road from scratch, it costs hundreds and hundreds of millions of dollars,” Marc Tarpenning told me. “Our whole business plan was based on the fact that we weren’t going to do that - we were going to develop a drivetrain and then we were going to take a design that had already passed the crash and safety tests both in Europe and America, that we knew could qualify as being a real car.”
“We quickly convinced ourselves that we could make a compelling electric drive. That’s Silicon Valley stuff - computers and batteries,” said Tarpenning. “But I was extremely concerned with our ability to actually make a car, because there are thousands of components and it just seemed like that was impossible. What I learned, by doing a bunch of research, was that the car industry had sort of re-factored itself starting in the 70s, and that they had essentially gotten rid of everything except a couple of core things. They kept marketing, some styling, and internal combustion engine development, although not the electronics and engine control units... almost all the [other] components come from other companies. So, that was what convinced me that this business was actually possible, because we could buy from those same suppliers.”
Above: Recently revised and updated book, Tesla: How Elon Musk and Company Made Electric Cars Cool, and Remade the Automotive and Energy Industries, edition 3.0 (Source: Charles Morris)
“I also discovered that the car industry had taken outsourcing to such an extreme that there were even companies that did the final assembly. Magna Steyr in Austria made 100% of the world’s BMW X3s, for example. I thought we could actually do this because we didn’t even need to have final assembly. We’d supply the drivetrains... then we would decide where to put the drivetrains into the gliders. A glider is a car that’s built without an engine and it’s commonly done because engines come from one factory and parts of the final assembly come from other factories and then they’re put together.”
In early 2005, Lotus provided Tesla with a couple of glider versions of its Elise two-seat roadster. When the boys installed their electrified innards into the glider, they had themselves a mule, which is a working car that can be driven to test various components on the road (according to Tarpenning, it’s called a mule because, like the equine animal, it doesn’t reproduce).
“We knew we could make an AC induction motor,” Tarpenning continued. “The battery system was something that was scary. We wanted to get rid of that risk as soon as we could. So we worked on the battery system first and we made a Lotus Elise into a mule, using our first-generation battery pack, an AC Propulsion inverter and an AC Propulsion motor. We drove that around, and that allowed us to get more people interested and to raise more money. It validated that the battery system was going to be okay, we were going to have a couple hundred miles of range and it was going to be really fun to drive. It was a great test platform for all kinds of stuff.”
Lotus provided the airbag system and ABS brake system, components that would have been very difficult for Tesla to develop on its own. This is why the Roadster’s steering wheel and dashboard look pretty much like the Elise’s. Most of the rest of the interior is all Tesla. The Californians also used the Elise’s windshield and other windows (called the greenhouse or glasshouse in the trade). To roll their own would have meant dealing with a host of related issues including visibility requirements, rollover protection, waterproofing, windshield wipers and more.
Above: A young Elon Musk examines the original Tesla Roadster (Image: Inverse)
However, the limits of outsourcing soon became apparent. After some contentious discussion, the Tesla team decided to use carbon fiber composite (aka carbon fiber reinforced plastic or CFRP) for the Roadster’s body, and once this decision was made, the team was no longer building a modified Elise, but a completely new car, so they had the freedom to design the body however they wanted. They invited several designers to submit concepts, and ended up choosing a design by Barney Hatt, a member of the Lotus Design Studio.
To say that the Roadster was “based on” the Lotus Elise would be a vast oversimplification. Tarpenning told me that the two cars share around six percent of their components. He also said that Tesla replaced the AC Propulsion drivetrain with a custom solution before production began.
In a 2012 interview, Elon Musk mused that the strategy of adapting an existing vehicle hadn’t worked out so well. “As it turned out, the AC Propulsion technology didn’t work, so we had to redo all that. And the Elise - once you added the electric powertrain, it invalidated all the crash work, the mass grew by 30 percent, the weight distribution was different, the load points were all different. We had to stretch the chassis just to be able to fit people in, so that turned out to be a really dumb strategy, too. So it was like you wanted to build a house, couldn’t find the right house, so you try to fix an existing house and end up changing everything except for one wall in the basement. It would have cost way less to just level the house and start from scratch. What sounded like a good idea at first, which was to leverage the Elise chassis, was actually an incredibly dumb idea.”