Three years ago, staid and practiced Toyota announced it would build a new city. A city that would test future technologies, that would be the birthplace of a new in-house startup called Woven Planet. A city that would show the world what the next generation of transportation would look like, where the best and brightest minds could collaborate to make cars safer, more efficient, and more autonomous.
Three years later, the city is a dirt lot full of cranes and girders. That fancy startup has been brought more tightly under steadfast Toyota’s control, and the release date for the software it was meant to deliver has been pushed back. What happened to Toyota’s dream for the future?
An investigation from the Wall Street Journal sought to answer that question. It tracked the rise and fall of Toyota’s software efforts, and how cultural mismatch turned big dreams into spreadsheets and timelines. From the Wall Street Journal:
The software unit’s early missteps, described by current and former executives, offer a cautionary tale for leaders of traditional companies who know they need a new kind of tech expertise but struggle to meld it with old aptitudes and habits. The Woven unit started with far-reaching ambitions that ultimately didn’t match Toyota’s desire for concrete software features ready to be put in cars fairly quickly. Woven struggled to connect with the changing goals of its parent, which is trying to reinvent itself while building 10 million vehicles a year.
The result was deadlines that kept getting moved forward and back—at one point, as far out as 2027 for the full rollout of the software. Today, Toyota has installed veterans of its auto-making group and brought Woven closer to the rest of the company. With its new regime and structure, Toyota and Woven say the unit’s plans are now back on track, with an early version of the software promised by 2025.
Every company says it wants a startup, but most seem to be falling victim to survivorship bias. CEOs look at these multimillion or multibillion-dollar companies that seem to have sprung out of nothing, and decide that implementing the same spirit will help their traditional company do more with less. What those CEOs don’t see is that most startups fail. Again from the Journal:
Woven Planet began using high salaries to attract top foreign software engineers and acquired a string of startups, including the self-driving unit of U.S. ride-hailing company Lyft in April 2021.
For its Silicon Valley-style office, with Segway-like personal transporters and plants to promote relaxation, Toyota took space in a central Tokyo high-rise building. It was 180 miles from headquarters in Toyota City, where the company has been known to use less air conditioning and shut down elevators to cut costs.
To run Woven Planet, Toyoda picked James Kuffner, a jeans-and-T-shirt-clad roboticist who had been with the carmaker since 2016, after serving as part of the team that created Google’s self-driving car.
This startup mentality was meant to help develop next-generation software for Toyota’s cars, a suite called Arene that would enable such futuristic innovations as over-the-air updates and data collection. Instead, the startup apparently failed to even figure out what the software would be:
Around six months ago, at an all-staff meeting at Woven, an employee’s three-word query—“What is Arene?”—was voted to the top of the list of questions posed to company management. There was a brief pause before the CEO, Kuffner, said he would answer.
Kuffner spoke at length about his original dream for Arene: saving lives with its automated safety functions and going beyond simply controlling the driving functions of a car. People who attended the meeting said they recall wondering why much of the answer was given in the past tense.
The full piece from the Wall Street Journal is worth the read. It paints a picture of a tale as old as time (or, at least, as old as modern investor-driven Silicon Valley valuations) — a traditional company wanting all the upsides of a startup with none of the flaws. But running a startup, on its own, is a massively difficult endeavor. Adding in the culture clash of trying to do so within an existing company — let alone one as traditional as Toyota — is a recipe for failure.