Wharton’s John Paul MacDuffie and Sharon Sandeen of Mitchell Hamline School of Law discuss the Google-Uber self-driving car lawsuit.
The race between companies to be the first to perfect and commercialize driverless cars detoured last week into a court battle between the most ambitious contenders, Google and Uber. Mountain View, Calif.-based Waymo, a subsidiary of Google’s parent company Alphabet, has accused the San Francisco-based ride-sharing service of stealing its designs and other proprietary information.
The route for that alleged theft is through Otto, a startup co-founded by Anthony Levandowski, a former Google executive who led the company’s self-driving car project and quit in January 2016. A few days later, he founded a startup called 280 Systems, later changed its name to Otto and sold it to Uber in August 2016 for $680 million. He now heads Uber’s self-driving cars project.
The first London Value Investor Conference was held in April 2012 and it has since grown to become the largest gathering of Value Investors in Europe, bringing together some of the best investors every year. At this year’s conference, held on May 19th, Simon Brewer, the former CIO of Morgan Stanley and Senior Adviser to Read More
“Otto and Uber have taken Waymo’s intellectual property so that they could avoid incurring the risk, time and expense of independently developing their own technology,” says Waymo in the lawsuit it filed on February 23 with the U.S. District Court in San Francisco. “Ultimately, this calculated theft reportedly netted Otto employees over half a billion dollars and allowed Uber to revive a stalled program, all at Waymo’s expense.”
“This is a big deal,” said Wharton management professor John Paul MacDuffie, who is also director of the school’s Program on Vehicle and Mobility Innovation. He noted that comparisons are being made between the Waymo case and the 2012 legal battle between Apple and Samsung over alleged patent infringement “in terms of the scope of the accusations and the implications.” (Apple won more than $1 billion in damages from Samsung in that case relating to iPhone design patents.)
According to Sharon Sandeen, professor of law and director of the Mitchell Hamline School of Law’s Intellectual Property Institute, the case could have either of two outcomes. One could be a settlement with cross licensing of the technology in question, where each company could benefit from the sharing of information, she said. “The other would be an all-out war to see who would win.”
MacDuffie and Sandeen discussed the implications of the Waymo suit on the Knowledge@Wharton show on Wharton Business Radio on SiriusXM channel 111. (Listen to the podcast at the top of this page.)
At the heart of driverless cars — and Google’s charges against Uber — is the so-called lidar technology, which in itself is not proprietary, but various companies have developed it further and patented their versions. The laser-based scanning and mapping technology creates real-time 3D images that allow a vehicle to “see” its surroundings and thereby allow a self-driving vehicle to detect traffic, pedestrians, bicyclists and any other obstacles a vehicle must be able to see to drive safely, the lawsuit stated. Google had refined its lidar technology to make it sufficiently cost-effective, thereby removing a critical barrier to commercializing the technology.
“[One outcome] would be an all-out war to see who would win.”–Sharon Sandeen
“[Lidar] is super expensive, and getting it to be both super accurate and cheaper is a huge race now,” MacDuffie said. Ford and Baidu have partnered to invest $150 million in a firm called Velodyne that makes lidar systems, while Israel-based Mobileye also makes them, he noted.
A Suspense Thriller
In its lawsuit, Waymo has built its case like a whodunit. Google’s investigation began after a Waymo employee was inadvertently copied on an e-mail from a supplier of lidar components to people who Waymo believes were working with Uber. Attached to the e-mail were drawings of Otto’s lidar circuit board, which Google says looked just like the Waymo design.
Waymo’s lawsuit further claims that six weeks before Levandowski quit Google, he had downloaded 14,000 confidential files containing trade secrets and patented designs for self-driving cars and attempted to remove any traces of that act. In November 2015, he had founded a company called 280 Systems that later became OttoMotto. In February 2016, he founded Otto Trucking, which Google at the time didn’t see as competition. Otto was formally launched in May 2016, and by July, it had lured two other senior executives who worked on Google’s self-driving cars project. Meanwhile, Levandowski received his “multi-million dollar” severance compensation from Google. By August 2016, he sold the two Otto startups to Uber.
Waymo has accused Uber and Otto of violation of trade secrets and patent infringements and sought an injunction to immediately and permanently bar the two companies from using the technology in question. “An injunction … would create huge problems for Uber,” MacDuffie said, because it would prevent the company from continuing to perfect the technology for driverless cars. Uber would also be hurt from “the public perception of it not continuing to be a lead innovator in this,” he added.
First In, or Out?
The pressure to be the first to commercialize safe self-driving cars is the main driver behind the actions of the various players in this case. MacDuffie said Uber bought Otto because it saw “an existential threat” if Google ended up dominating the self-driving car market, referring to an interview in Business Insider magazine in August 2016 with Uber founder and CEO Travis Kalanick.
“If we are not tied for first, then the person who is in first, or the entity that’s in first, then rolls out a ride-sharing network that is far cheaper or far higher-quality than Uber’s, then Uber is no longer a thing,” Kalanick said in that interview. Waymo in its lawsuit has cited this interview as a piece of evidence to support its accusation against Uber and Otto. “Uber felt the need to act swiftly to harness that technology and make it work for its business model,” said MacDuffie.
Google had been a first mover with self-driving cars, launching its project in 2009 and making it public in 2010 with statement that “Larry [Page] and Sergey [Brin] founded Google because they wanted to help solve really big problems using technology.” By 2014, the company had developed its own “reference vehicle,” a two-door autonomous car without pedals or a steering wheel. A year later, a prototype made a road trip in normal traffic on public roads. In 2016, Google moved its self-driving car project to Waymo, created as a subsidiary of the parent Alphabet. Waymo has thus far logged more than 2.5 million miles in autonomous mode on public roads. Google, or Waymo, has conducted trials with self-driving cars “for a longer period than any other company in that space,” said MacDuffie.
“Given the financial stakes, it seems like Uber would want to settle.”–John Paul MacDuffie
MacDuffie noted that Levandowski was a key early Google employee in its software development for autonomous vehicles. But speed-to-market was a factor that probably drove him to leave Google and join Uber. When Uber bought Otto, Levandowski was reportedly frustrated “that Google was slow in deciding to take the big step in terms of commercializing the technology,” MacDuffie said.
Meanwhile, MacDuffie said many “unknowns” exist today about driverless technology. He noted that a proliferation of different systems that operate differently will make it harder to achieve the goals of reducing accidents and deaths, and reaping other savings. Standards for driverless technology could come about in one of two ways, he added: Either regulators could lay down standards that are validated by tests, or competitively some company could secure such a lead to become the de facto standard.
MacDuffie pointed out that many aspects of driverless technology are currently unregulated. Some states have formed their own rules on permitting trials on their state roads, while the federal government is taking a wait-and-see approach even as it has issued some broad guidelines, he said.
Undeterred by that lack of regulatory clarity, Uber piloted its self-driving technology in Pittsburgh in September 2016. The company went ahead with some support from the city to conduct that experiment, but without any other specific legislative approval, said MacDuffie. “[Uber] has tended to do that in general — not worry whether what they are doing is legal, but jump in and prove it and figure they can deal with the consequences later.”
The Legal Terrain
MacDuffie suspected that the effort in the Waymo lawsuit is “to create a story of some nefarious intent.” For example, Waymo has said Levandowski was meeting with Uber executives even before he left Google to found 280 Systems. Otto Trucking and OttoMotto were “in stealth mode for several months,” it noted. “Google will argue that going to work for Uber was Levandowski’s plan along, and the other activity he was involved in were deceptions,” said MacDuffie.
“In trade secret cases, when you have evidence of [the] nefarious behavior the complaint alleges, you can create a picture that there is something wrong here — that has a lot of power,” added Sandeen. The onus is on the defendant’s attorney to point out that all or most of what Levandowski took was gathered by him before he joined Google, she added.
As it happens, Levandowski was working on driverless technology before he joined Google. After he completed his bachelor’s and master’s degrees in industrial engineering and operations research from UC Berkeley, he had started working on technology for autonomous vehicles “going back to at least 2004,” Sandeen noted. He later founded a startup called 510 Systems that he sold to Google in 2011 and joined the company at the same time. “There is at least a great body of information that [Levandowski] probably [already] had” before he joined Google, she said. In setting up the Otto companies after leaving Google, “maybe [Levandowski] was just doing that again,” she added.
“[Lidar] is super expensive, and getting it to be both super accurate and cheaper is a huge race now.”–John Paul MacDuffie
According to Sandeen, most trade secret cases involve former employees going to work for competitors or leaving to create a startup. Under trade secret law, “employees get to keep their general skill and knowledge earned over time.” She said the questions in the latest case are not only about whether or not Levandowski took information away with him, but what the nature of that information was.
Sandeen suggested that some of that information might be related to work that he had done before joining Google or research at UC Berkeley. “That will be a key inquiry in the litigation,” she said. It is also not clear if Google obtained a non-compete agreement with Levandowski when he left, or if such an agreement had expired at some point in time, she added. “Under trade secret law, independent development of the same technology is perfectly allowed.”
In terms of how the lawsuit might proceed, MacDuffie said the challenge is getting hard evidence of the theft of proprietary technology. The misdirected email may be damning, he noted. “That piece of evidence does help the strength of [Google’s] case,” he said.
Sandeen said she found it “curious [that Google] waited so long” to act after Levandowski quit. “Normally — at least [according to] the best practices that are developing now — that’s the point at which you would have computer forensic experts come in to determine if anything had been taken,” she said. But the lawsuit indicates that Google didn’t get suspicious until months later.
Sandeen noted that the best practices for trade secrets require entrance interviews to prevent having any material coming in that would taint the R&D work being done at the employee’s new workplace. Exit interviews are also recommended to make sure no proprietary data was being taken away, she said. Google may have followed that process, but perhaps representations were made about the nature of the information that didn’t cause any alarms at that time, she added.
“Under trade secret law, independent development of the same technology is perfectly allowed.”–Sharon Sandeen
Troubling Times for Uber
MacDuffie recalled that Uber and Google were closely aligned in earlier years. Google Ventures had invested $250 million in Uber in 2013, and placed a senior Alphabet executive, David Drummond, on Uber’s board. Google Maps were tweaked to offer Uber as an option for users, with discounts thrown in. Google was “hugely helpful” in Uber’s growing its customer base, he said.
However, the recent developments have strained that relationship, MacDuffie noted. Drummond resigned from Uber’s board in August, soon after the conflicts between the two companies became known. In September 2016, Google announced a new car-sharing service would become part of its Waze navigation app, and observers suggested that it could be the first step in launching a self-driving car service that would compete with Uber. “That’s a head-on competitor, with a slightly different twist, to Uber and Lyft and the like,” he said.
The Waymo lawsuit couldn’t have come at a worse time for Uber. MacDuffie noted that Uber has been getting some bad press of late. “Accusations that it is stealing secrets from another company — whether that is true or not — will fit in a little bit with their public image of being brash and aggressive and not worrying about rules,” he said.
Two weeks ago, Uber launched an investigation into sexual harassment claims by a former employee. Last week, in an unrelated move, it let go a senior executive for allegedly failing to disclose that he had faced sexual harassment charges at Google, where he had worked earlier. Uber’s business model has also come under intense scrutiny as it burns cash, as a recent Knowledge@Wharton article noted. After its pilot of self-driving cars in Pittsburgh last September, Uber launched pilots in San Francisco in December. But that lasted only about a week before California officials ordered Uber to stop the autonomous car service, which officials termed as illegal, citing traffic violations and registration lapses by the vehicles involved.
Under those circumstances, MacDuffie said Uber would hate any injunction that stops it from learning more about driverless technology. He expected that a settlement would be very much in Uber’s interests, sooner rather than later. Uber has said it would fight the case strongly, he noted. “[However,] given the financial stakes, it seems like Uber would want to settle.”
Article by Knowledge@Wharton