Originally published on the EV Annex blog.
By Charles Morris
One of Tesla’s big advantages over legacy automakers is its software expertise. Back in 2014, Tesla co-founder Ian Wright told me that Big Auto’s software tended to resemble “a dog’s breakfast.” Tesla, with its Silicon Valley roots, designed its vehicles with a unified systems architecture from the beginning. “That’s the sort of thing that Tesla gets right in their sleep, and the big guys really struggle with,” said Wright.
|FANUC Robot Assembly Demo for Elon Musk at Tesla (YouTube video by Steve Jurvetson)|
Half a decade later, they’re still struggling, as VW’s software snafu with its new ID.3 EV demonstrates. However, there are some deeper lessons to be learned from Tesla’s software mindset.
Steve Denning is an authority on Agile management, a methodology that prioritizes responding to customers’ needs and embracing change. In a recent article in Forbes, he argues that hardware developers need to embrace Agile principles to keep up with the pace of innovation.
The Agile Manifesto, created by 17 software developers in 2001, embraces four core values: individual interactions over process and tools; working software rather than thorough documentation; collaboration with customers; and responding to change rather than following a plan. For Denning, the primary principle of Agility is “an obsession with the customer’s needs.” He writes that “innovation begins from, and is driven by, the customer’s changing needs.”
This is reminiscent of Steve Jobs’s famous line: “Start with the customer experience and work backwards to the technology,” a principle that Tesla has followed to great success. But what does this have to do with the dichotomy between hardware and software?
Plenty. Many have pointed out the obvious fact that atoms are harder to manipulate than ones and zeroes, and have cited this as the reason that innovation tends to happen much more slowly in the physical realm. I myself have often drawn comparisons between the late-1990s internet boom and the current transition to electric mobility. The latter is playing out much more slowly, and it’s tempting to assume that this is only because building automobiles (“metal-bending”) happens in the physical realm. As I’ve quipped in more than one previous article, cars, unlike software, can’t be built in a dorm room.
However, Denning argues that, “the atoms-vs-bytes difference is far from being the whole story. Existing manufacturing processes are also slow because they are designed to be slow.” He presents the example of a car door: engineers may come up with a better design, but if an automaker has $100 million invested in a machine that makes the existing door, the innovative new door isn’t likely to be put into production any time soon. Today, with 3D printing techniques, it’s possible to make rapid changes to hardware components at a fairly low cost. Why doesn’t this happen? It’s not a hardware problem, Denning tells us, but a management problem.
A look at how Tesla’s share price compares over a five-year time period with other OEMs (Image by Steve Denning)
In the pre-Tesla days, you bought a car from a dealer’s lot, and that was the car you had until you wore it out or sold it. Buyers didn’t expect anything else from a car. However, consumers who’ve grown up with computers and smartphones are “coming to expect what Tesla already delivers: a capability to improve their cars every month or so after purchase through software updates,” writes Denning.
Agile software development is all about designing products in a modular fashion, continually innovating and incorporating customer feedback and lessons learned in the development process. Agile developers “never arrive at perfect,” writes Denning. “They keep at it. The improvement journey never ends.”
Elon Musk has brought some of this continuous-improvement mojo to the hardware world. “Tesla [uses] traditional CAD tools such as CATIA, but they also invested in an end-to-end 3D modeling system from which they can view and simulate entire assemblies and automatically print parts,” consultant Cliff Berg writes. “Importantly, the software is fast, even when handling complex assemblies, so that engineers do not have to wait, which encourages a rapid iterative design approach.”
SpaceX also exhibits a similar process of continuous iteration. “Instead of spending months or years on design and then carefully building one perfect prototype, they build many, and they test them all in myriad ways,” writes Berg. “They want to get the thing to the test stand, even if it has imperfections. To reduce the cost of making parts whose design is always changing, they use 3D printing. This … allows the design to rapidly evolve.”
Musk’s two main companies, Tesla and SpaceX, are case studies in a new way of looking at manufacturing. As Denning writes, “they demonstrate that it is possible to run a hardware company like a software company in an Agile fashion.” Could a legacy automaker (or aerospace firm) make the leap to this new way of thinking?
As many a management guru has lamented, traditional ways of doing things are embedded deeply in the majors’ corporate cultures, and so are outdated definitions of their products. “Car companies are still thinking of a car as a metal transportation device with a few electronic gadgets attached to it,” writes Denning, echoing what Ian Wright told me years ago. The central truth about the slow pace of automotive innovation: “It’s a leadership and management problem, not a problem of atoms versus bytes.”