The fighter for justice did not give Waymo to patent the key technology of lidar

Eric Svildens has no other personal interests in this matter, except intellectual curiosity.

The lonely engineer succeeded in what the best lawyers and expert witnesses from Uber could not achieve — stop the registration of the patent covering the lidar technology (laser range finders) from the main rival, Waymo.

Working on an unusual complaint filed by Eric Svildens, the US Patent Office rejected almost all but three out of 56 applications from Waymo patent number 936 (according to the last three digits of its serial number). The bureau found that some of the statements reproduced the technology described in the Velodyne lidar manufacturer's patent, while others were simply "impossible" and "akin to magic."

Svildens, who does not receive any money or benefits from this decision, told the editors of Ars magazine that he was terribly pleased with the news. “This patent wasn’t worth filing at all,” he said. “It is well written, but I believe that this is the case when they claim the invention of what they did not invent.”

Patent 936 played a key role in last year’s epic lawsuit against intellectual property and Uber. In December 2016, a Waymo engineer accidentally received a copy of an email sent by one of the suppliers to Uber, which contained a lidar circuit that was almost identical to that indicated in patent 936. The

patent describes how to set up a laser diode to emit pulses using a circuit that includes inductor and gallium nitride transistor (GaN). This accidental discovery helped file a lawsuit in which Waymo accused Uber of infringing patents and using the secrets of lidar allegedly stolen by engineer Anthony Lewandowski .

In August 2017, Uber agreed to redo the Fuji lidar so as not to violate patent 936. Then, in February 2018, Waymo settled problems with industrial espionage in exchange for a share in Uber worth $ 245 million and the company's promise not to copy their technology. “An agreement was also signed to ensure that Waymo’s confidential information will not be used in Uber, including hardware and software,” a Waymo spokesman said at the time.

Now it turns out that this change in the scheme was unnecessary, says Svildens, an engineer who asked the patent bureau to become more familiar with the patent 936. ", - he said.

Uber told the editorial board that, despite the court’s decision, it will not change the design of the lidar again.

Just an interested observer

Interestingly, Svildens does not work in Uber, or in Velodyne, or in any other company that develops a mobile. He works in a small startup related to cloud computing. Svildens became interested in a patent that came up in the Uber case, and he saw how simple the Waymo lidar scheme seemed. “I could not believe that this scheme did not exist prior to the registration of this patent, ” he told Wired magazine last year.

Svildens study revealed the existence of several patents and books published before the patent was registered by Waymo. He spent $ 6,000 of his own money on organizing a formal protest against the patent. Waymo responded by sending dozens of documents, calling experts as witnesses, trying to remake several applications and diagrams from a patent to ensure its survival.

But she did not convince the patent office. In March, the bureau expert noted that the redrawn diagram of the Waymo lidar start-up showed that current flows along the wire between the circuit and the ground in two directions, which is usually considered impossible. “The testimony of the witness on the part of the patent owner could not convincingly prove that the circuit is generally grounded, taking into account the presence of a magical ground wire, the current through which moves in two directions,” the expert noted dryly.

In September, the case finally resolved. With some exceptions, it was found that those applications that were not physically impossible copied the lidar schemes existing in previous patents, in particular, in patent 558, issued to Dave Hall, founder of the lidar manufacturer Velodyne. Svildens says that the result does not surprise him, and believes that it was not for nothing that he spent his money. “When I studied patent 936, it became clear to me that it was incorrect due to the presence of previous works,” he says. “I applied for re-examination only because I was absolutely sure the patent was invalid.”

Waymo declined to comment, but now the company is in an awkward position. In 2017, Waymo Director John Krafchik announced publicly at the Automobili-D Automotive Conference in Detroit that the company, which had previously bought dozens of lidars from Velodyne, now develops and builds all its sensors independently, and from scratch.

Then Krafchik continued to pin up his former supplier. “Developing our own lidars not only gives us a more reliable product than the purchased versions, but also allows us to get it much cheaper,” he said. “Just a few years ago, one lidar of maximum range [from Velodyne] cost $ 75,000. Today we have reduced its value by more than 90%. ”

Svildens believes that the discoveries of the patent office put these statements in question. “The Waymo lidar scheme seems to be based on the Velodyne lidar scheme used by them,” he says. “In addition to using the Waymo scheme, which is similar to the Velodyne scheme, the company also tried to patent (and patented) these schemes as its own inventions.”

Look who's Talking

In its first lawsuit against Uber, Waymo wrote: “Waymo developed its patented inventions, spending significant funds on it and years of hard and hard work, experimentation, trial and error. If defendants are not prohibited from violating intellectual property rights and illegal possession, this will cause Waymo to suffer serious and irreparable damage. ” If Velodyne considers that their patent 558 covers the lidars that Waymo is about to market, they will be able to file a similar complaint.

However, the situation is complicated by the fact that Velodyne is currently stuck in her own patent dispute with Quanergy, a startup developing solid-state lidars. Velodyne accused Quanergy of infringing intellectual property rights, and Quanergy responded by declaring the patent 558 from Velodyne.

Velodyne is trying to support the patent 558 by rewriting it, but the judges in this case noted in a recently filed note that there is a “reasonable likelihood” that at least one of the company’s statements is not patentable. Velodyne refused to comment to the magazine on the case.

Robo-mobile startups should not accept this legal confusion as a call for using the lidar technology described in Waymo and Velodyne patents, warns Brian Love, one of the directors of the Institute of High Technology Laws at the Law School in Santa Clara. “Patent lawyers have a joke that a final failure is never final, because owners always have room for maneuver even after a final failure,” he told the editorial board. - And to win in a patent court, you only need to prove the only violation in a single patent. And the fact that there is only one statement left in the Waymo patent indicates that there is one chance to assert that someone has violated this application. ”

Waymo has hundreds more patents.developed in the company or acquired from competitors. Not all of these patents may contain “magical” statements of the 936th, but they will probably be enough for the company to remain at the helm in the area of ​​work mobiles for some time - especially due to the fact that, according to Svildens, Too busy to submit any other patent claims.