Monday, November 20, 2006

Tesla Motors

You've heard about Tesla Motors already--the silicon valley startup that's making the next generation in electric cars, a roadster that can go from 0 to 60 in 4 seconds, and that looks like a million bucks. (It'll also cost nearly $100K, but that's cheap compared to a Lamborghini, and you just might beat it off the line.) There have been plenty of great articles on the company and the car, from Wired, The Guardian, and others. As it happens, I know several of their engineers, so I was able to get a tour of the company a few weeks ago. This wasn't an official interview with company managers or spokespeople, so I didn't get answers to all of my questions, but my friends Drew and Colin knew most of the answers, and I was also able to squeeze in a couple minutes with JB Straubel, the company's Chief Technical Officer. Here's a summary of their answers to some technical questions that other news sources haven't written about. (The insider geek's view of Tesla Motors, if you will.)

Their building is practically hidden in San Carlos, an unobtrusive light-industrial space sitting off the beaten path amongst warehouses and more blue-collar industry than most of Silicon Valley's sprawling office parks contain. But once you step inside, the cover is blown, and you can tell there's something exciting happening. You can also tell they're a high-tech company, not a normal car company. They're small, and dominated by engineers. In fact, they're effectively just an engineering company, since both the aesthetic design and the manufacturing are outsourced to Lotus. (That's why the Tesla looks like an Elise.)

Aerodynamically the car is good, but its body is designed for looks, first and foremost. It has a drag coefficient of around .3, as compared to the EV1's .19. So although the Tesla gets 250 miles on a charge--which is excellent, more than double most EV's--the range could have been far longer. Some EV-geek chatter at has suggested by back-of-the-envelope calculations that the power train of the Tesla in the body of an EV1 or an Opel Eco Speedster would have a 400 mile range. Conversely, if they had designed the car with the EV1's drag coefficient and kept a 250-mile range, they could have eliminated a lot of the battery bank, and thus a fair chunk of the cost of the vehicle. However, I don't begrudge them their decision to prioritize looks over efficiency. The Tesla, being electric, is already far more efficient than a combustion-engine car. It's a big step in the right direction, and a 250-mile range at a $100K price tag for the first run of an amazingly hot all-electric roadster is great. Too often we let the perfect be the enemy of the good.

The hardest part of an electric vehicle is not the motor, it's the controller--the power electronics which drive the motor, handle regenerative braking, and manage the power from the batteries. Tesla chose AC Propulsion's controller because of their expertise and reputation as the best in the business. Their founder and lead engineer, Alan Cocconi, is apparently so smart that instead of having CAD-refined schematics and models of the circuitry, he had the whole layout in his head. Before designing the controller for AC Propulsion, he designed the controller for the GM Impact (which became the EV1).

Tesla's motor is a high-performance induction motor; not revolutionary new technology, but top-of-the-line. In fact, it's the same one the EV1 and AC Propulsion's tZero prototype cars used. Tesla's innovation is in the way it is manufactured, keeping performance quality high but reducing costs. I also asked whether they thought about using in-wheel motors, since putting a small motor in every wheel instead of having one big motor with a drivetrain connecting it to the four wheels can greatly reduce mechanical complexity and weight, as well as improving reliability. (This is one thing EV's make possible which simply can't be done feasibly with combustion engines.) Interestingly, they did consider it, but JB said it would have made safety certification extremely difficult. It's perfectly safe, but the certification regulations are written assuming you have one motor and a drivetrain, so there are some certifications (such as the one for Anti-Lock Braking) you can't pass in a car with no drivetrain. These rules would need to be re-written to allow vehicles with in-wheel motors to be certified, which is obviously not going to happen without significant money and time spent lobbying--not a fight a small startup company should take on if it can avoid it.

The company has been operating for several years in "stealth mode" as many startups do, but started making a splash in the press this year despite the fact that their cars won't actually be available for another year. Why come out of stealth mode now? Since my friends are engineers, not the business folks, their answers were speculative, but the investors are silicon valley people, and the way things are done out there is generally to start creating the buzz before the product is out. Begin in stealth mode when you're not really sure whether it will work or how long it will take, but once you're reasonably assured of your technology and its prospects for success, let the whole world know so you can build anticipation. Tesla certainly knows that their technology and design work--they have had working cars for a long time now, and are just going through the industry-standard process of testing (a multi-step, multi-year process, from the sound of it) that works out the bugs in manufacturing and such. It will be exciting to see what kind of splash their cars make when they officially release them next year. It will also be interesting to see what their sedan will be like--they don't have any built (or even fully designed) yet, but after their roadster is a big success on the market they plan to expand down the food chain to vehicles for the rest of us. The sedans still won't be cheap, but they won't have Ferrari price tags.

My last question was, how will Tesla succeed where other EV companies have failed? The last two decades are littered with the wreckage of failed ventures in electric vehicles: Corbin Motors's Sparrow, Solectria's Force, GM's EV1, Ford's Th!nk, to name a few. Some of these companies had engineering problems, some of them were trying so hard to make their vehicles affordable that they couldn't make a profit, and for GM, in addition to technical hurdles, the company executives apparently felt that it wasn't in their interests to succeed. Tesla will be different for a few reasons. They've worked with the realities of the market--EV's require expensive technology, so you might as well accept it and make a car that people are happy to pay lots of money for because it's amazing, rather than making lots of compromises and requiring your customers to compromise as well. This means they have the money to hire top-notch engineers and work out all the bugs properly. Being expensive but high-quality in the beginning is also a smart road to future affordability, because reducing costs is easiest to do on a known problem, an existing product; it's much harder to make something both cheap and good right out of the starting gate. (Also, as they pointed out in their Forbes interview, it is much easier to move down in a market than move up in one.) Tesla is also not trying to do too much themselves, with the experts at Lotus able to handle many of the problems that would hamstring an auto startup doing their own manufacturing. (Though they have hired away many Lotus engineers for themselves, too.) They understand the march of advancing technology, while the big American car companies don't. And also unlike existing car companies, Tesla is unencumbered by chains of vested interests, they have the will to succeed.

Original article