GM Quietly Deepens Silicon Valley Ties By Learning To Play Tech Incubator – Forbes
Early this month a handful of high-level engineers from General Motors’ R&D division joined Silicon Valley venture funds in a packed Mountain View auditorium to hear dozens of two-minute pitches from tiny tech startups eager for early-stage funding. The GM crew wasn’t there to invest in or buy any of those companies, at least not directly, but to identify a few promising ones to work with as a mentor.
The automotive giant is backing 500 Startups’ Demo Days in a partnership with the tech incubator forged late last year. GM, like every other major automaker, already has an advanced tech unit in Silicon Valley, where its scientists and technology specialists in Palo Alto beat the bushes for new ideas and companies.
But the relationship with 500 puts “scouting on steroids – and we’re learning a lot about startup culture too,” Gary Smyth, executive director of GM’s Global R&D Laboratories, told Forbes.
It formed that relationship prior to reports of sexual harassment by 500 co-founder and chief Dave McClure, who resigned in June. Christine Tsai, also a co-founder and 500’s new leader, was frank about the damage done by McClure and committed the firm to being part of the solution to a wave of reports of inappropriate executive behavior roiling Silicon Valley. 500’s “mission is much bigger than just one person,” Tsai told the Demo Day audience. “It’s way too important to be taken down by one person’s mistakes.”
Certainly GM hopes so because the biggest U.S. automaker expects the firm to act as something of a Sherpa, helping it learn to work effectively with creative but unproven new firms.
“These early stage companies are very fragile. You’re watering them, you’re feeding them, but you have to be careful and let them grow,” said Smith, a burly veteran engineer hailing from Northern Ireland. GM can assess technical capabilities, but is less skilled at assessing founders of early-stage startups.
“You can look at the technology and look at the market that technology will go in and say `do they have something that’s unique, that isn’t me-too?’” he said. “What we’ve learned is the big factor is the founders – will they be successful? 500 has a strong capability of not just finding (companies) but looking at the founders of companies very differently.”
Listening to startup pitches is a quieter step than GM’s big splash in early 2016. That’s when it bought San Francisco-based autonomous vehicle startup Cruise Automation in a deal worth at least $581 million. And even as Cruise CEO and co-founder Kyle Vogt leads a rapid expansion of the unit to ensure Detroit-based GM is at the forefront of self-driving tech, the carmaker still has a big appetite for innovative technology and services.
GM won’t say how much it’s spending to back the 500 pitch sessions, though the amount is likely modest for a company of its size. The more important goal is learning to align its century-old engineering and research chops with flexible and fast-moving startups to create disruptive and revenue-generating ideas.
In past decades, advanced research done by armies of GM engineers at its Warren, Michigan, tech center might be interesting but could not always be commercialized, Smyth said. Since 2012, there’s been a sea change in how the company thinks about R&D, and that’s only growing more intense as Silicon Valley firms from Google to Apple to Tesla and Uber become ever bigger forces in the auto world.
Smyth’s message to GM researchers is simple: “Think how can I get game-changing technologies commercialized? It has to be high impact, not incremental, but more radical, disruptive technologies.”
On this hot August day, the companies that caught GM’s attention didn’t have radical automotive AI or powertrain breakthroughs – and weren’t even from Silicon Valley. But a few had ideas deemed useful to the company’s evolution.
They include Factory Four, a Baltimore company with ideas to improve 3D-printing-based production, and VR Motion, a Portland, Oregon, firm, using virtual reality to create highly detailed, vehicle training simulators that are of great interest to the military and police departments.
A 500 event earlier this year also led GM to work with UrbanLogiq, a Vancouver firm that harnesses real-time traffic and road condition data, including some from GM itself, to assist urban planning decisions. That session also brought Preteckt to its attention. Using sensors on commercial trucks, the Memphis-based company uses cloud-based analysis and machine learning to provide vehicle “prognostics” (rather than diagnostics) – early warnings of components determined to be at risk of failure days before a breakdown.