Tesla’s Next Broken Promise
Automakers from Shanghai to Detroit have responded to Tesla’s remarkable success by accelerating their own electric vehicle ambitions. Now Tesla is promising to upend large-scale car manufacturing, but it faces tough competition.
Tesla Motors and its chief executive, Elon Musk, have made an indelible mark on the global automotive industry with their celebrated battery-powered cars, purchased by consumers whose passion for Tesla appears boundless. If that feat just a few years ago seemed unlikely, the youthful company’s next exploit looks even more so: to prove that vehicles in the digital age can be manufactured at great speed and high quality with minimal human manual labor.
Since it started making cars in 2008, Tesla has delivered more than 140,000 electric vehicles globally, most of them its luxury Model S sedan and the Model X crossover with gullwing doors. Earlier this year, 400,000 people each sent Tesla $1,000 deposits toward its next car, the Model 3, scheduled to begin production next July in Tesla’s factory in Fremont, California. Aimed at domination of the mass market for electric vehicles, the Model 3 is a popularly priced, battery-powered sedan that will sell for as little as $35,000. (Nissan has sold 350,000 of the current market leader, the Leaf.)
With the Model 3, Tesla is blazing toward uncharted territory and doing so with characteristic brashness. To build cars for middle-class consumers, in far greater volume and more quickly, Tesla is designing a new type of factory. If it works, it would embody revolutionary innovation in mass manufacturing.
Tesla’s plans have not been spelled out in detail. The company didn’t furnish interviews for this story, but did answer questions via e-mail. Musk has previously declared his intent to “build a machine that builds the machine,” a highly automated vehicle factory that will stun conventional automotive manufacturers, one in which human workers will all but vanish, replaced by highly automated software-directed machinery and robots that will build Model 3s, and perhaps other Tesla models, at speeds greater than any accomplished in assembly plants operated by General Motors, Ford, Toyota, or any other global automaker. During the company’s second-quarter earnings call with financial analysts, Musk nicknamed his proposed factory “alien dreadnought”—a Star Wars battle cruiser from another galaxy.
“The point at which that's what the factory looks like, that's when you know you've won,” Musk said this summer. The futuristic assembly plant will unfold in three stages, while the number of people needed for assembly work will diminish, he said. “You can't have people in the production line itself, otherwise you drop to people speed,” Musk said. Remaining workers will be in charge of maintaining and upgrading machines, as well as “fixing anomalies.”
Musk faces new competition from automakers from Shanghai to Detroit, who have responded to Tesla’s success by accelerating electric vehicle ambitions of their own. GM is the farthest along and will begin production of the Chevrolet Bolt (to be sold as the Opel Ampera in Europe), its competitor to Tesla’s Model 3, later this year.
The way the GM cars will be built contrasts in important ways with Tesla’s method. Though crammed with modern tools, GM’s manufacturing system is stress-tested, the culmination of decades of incremental improvement designed to ensure high quality while meeting consumer demand—with little risk that the flow of vehicles to retailers and consumers will be interrupted or delayed.
By contrast, Tesla has consistently failed to live up to production forecasts. Now Musk, assuming that Tesla will be able to develop and debug a factory production system that operates unlike any the world has seen, has promised a jaw-dropper: that Tesla will deliver 500,000 vehicles in 2018, two years more quickly than an earlier forecast. Once that goal is achieved, Tesla’s output will grow by 50 percent annually, Musk says, reaching about a million vehicles delivered annually by 2020. Although that output would constitute roughly a tenth of the number of vehicles currently manufactured annually by GM, Volkswagen, or Toyota, it would elevate Tesla from a plucky niche producer to a serious industry player.
It will be a big jump for Tesla. In the first half of 2015, the carmaker built and delivered 21,537 vehicles from its single assembly plant in Fremont, California. That’s about half the total vehicles GM builds in one day at its 75 factories around the globe.
Musk’s grand manufacturing vision began with Tesla’s purchase of the Fremont assembly plant from Toyota in 2010. The CEO explained to analysts at the time that he was no more interested in replicating traditional automotive manufacturing processes than he was in designing traditional gasoline-powered cars. Accordingly, he recruited a number of manufacturing top guns from other industries, such as electronics.
Following several manufacturing delays, Musk conceded that Tesla’s designs, such as the troublesome gullwing doors on the Model X, might have been too complicated to build according to promised delivery dates. The automaker missed publicly announced production deadlines for its first three electric vehicles: the Roadster (built on a Lotus body), Model S, and Model X. In May, Tesla confirmed the departure of two top manufacturing vice presidents, one from the solar panel industry and another from an electronics manufacturer. Musk then issued a public invitation to global automotive manufacturing veterans to join the company. A week later, Tesla disclosed the hiring of Peter Hochholdinger, a veteran from Volkswagen’s Audi luxury subsidiary, to head manufacturing.
Even the company’s fan base and supporters wonder whether Hochholdinger can break the trend and produce cars, notably the Model 3, according to promised delivery dates.
“Tesla Motors operates on Tesla time. Quite frequently, the infectious optimism that pervades the company causes its leadership to announce release dates for products that are … well … aggressive,” writes Roger S. Pressman, an admirer of the company, its cars, and of Musk, and whose Deerfield Beach, Florida-based supply company, EVannex, manufactures accessories for Tesla vehicles. “Often, Tesla will miss those dates, sometimes by a little, sometimes by a lot.” Pressman has written a book called Getting Ready for Model 3.
A GM rival
One reason for that may be that Tesla’s manufacturing protocol diverges from the norm for mass manufacturers in the sourcing of parts and components. Automakers that serve the mass market tend to enlist outside manufacturers to produce parts for the sake of efficiency and quality, and to take advantage of supplier innovations. Tesla instead builds more of its own parts in-house, maintaining a higher degree of vertical integration and control than is typical for a mass manufacturer. Its “gigafactory,” currently under construction in Nevada, will supply all the batteries for its cars, for example. Electric drive motors and seats are two other components that Tesla builds within its own factory walls, but that most other manufacturers outsource to suppliers.
Contemporary automakers don’t publicly embrace Musk’s worker-free vision, yet their best plants are, in fact, highly automated, equipped with robots and computer-guided machinery. The number of workers needed for assembly-line production by GM and others has continued to fall, dropping by a third on average over the past 10 to 20 years. Supplier plants, which furnish the 10,000 or more components that typically make up a vehicle’s structure, are increasingly automated as well. (A few processes, such as the manufacture of wiring harnesses that form the backbone of a vehicle’s electrical system or the sewing of upholstery on seats require relatively more labor and thus are typically outsourced to lower-wage countries.)
This system is so well developed that suppliers to major automakers continue to specialize, innovate, and fragment into “tiers”; tier one suppliers make and send parts directly to final car assembly plants while tiers two, three, and so forth furnish components and materials to tier ones and to one another.
GM’s rival to the Model 3, its Chevy Bolt electric vehicle, will start production later this year in Orion Township, Michigan. The Bolt will be built on the same assembly line as two conventional gas-powered subcompact cars. The Bolt’s unique body architecture is designed for “manufacturability,” GM says, that is, specifically to accommodate the battery, electric motors, systems module, and cooling unit needed by an electric vehicle. Though an electric car lacks an engine, transmission, and other components needed for operation of an internal combustion engine, GM designed the Bolt so it can be assembled efficiently, at a high rate of production and at top quality, says Bill Shaw, a senior GM manufacturing executive.
Because the Bolt will be made on the same line as gas-powered cars, GM will be able to easily scale production to match market demand. “We do know we have to be agile. We’ve got a lot of levers to pull,” says Shaw.
1,000 cars a day
The goal, says Shaw, was to make the manufacturing of this vehicle as much as possible like that of other conventional models. Because the Bolt battery was designed to be installed where the gas tank goes on one of its conventional models, “the same workers can perform the two different tasks,” he says. This design minimizes cost by standardizing work and training, putting workers in a better position to avoid accidents, and ensuring that defects are avoided, says Shaw.
GM continues to add to its own high-tech manufacturing tools. Shaw described a computerized torque wrench recently developed that turns the fasteners securing Bolt’s battery to an exact tightness, while ensuring that none of the fasteners are misaligned or omitted. In case of an error, a computer notifies the operator.
The fantasy of a “lights out” plant—an automated plant so lacking in human presence that it could operate in the dark—has long been discussed in theoretical terms. In the 1980s Roger Smith, the late GM chief executive, pushed his manufacturing executives to begin to “robotize” assembly plants. Just as influential has been Toyota’s manufacturing system, which relies on low-tech principles such as just-in-time supply chain, standardized work procedures, and employee involvement in continuous improvement and efficiency to dramatically drive down manufacturing costs. GM developed its own version, the Global Manufacturing System, and for a time operated a collaboration on manufacturing methods with Toyota in the same building in Fremont that Tesla operates today.
State-of-the-art car factories rely on assembly-line workers as well as programmable machines. With a typical assembly line producing around 60 vehicles an hour or roughly 1,000 a day on two eight-hour work shifts, a line might, at the outside, be able to manufacture more than 250,000 vehicles a year. (Several North American automotive assembly plants operate on three shifts of six-and-a-half hours. The biggest produce more than 500,000 vehicles a year.) The two-shift total assumes production goes flawlessly. Factor in supplier parts shortages, machinery breakdowns, quality problems, labor issues and changes to vehicle design, and annual production drops to the more typical 100,000 to 150,000 vehicles per assembly line per year.
Those numbers show how hard a task Musk set for his company when he promised to ramp up production to 500,000 cars two years from now. The CEO may yet come up with some new surprise to keep his company in the game, but this deadline is destined to be missed.
Become an MIT Technology Review Insider for in-depth analysis and unparalleled perspective.Subscribe today