“You have to do it right the first time,” said Tekedra Mawakana.
Two men died in a crash in Texas, and authorities at the location said the Tesla Model S had no one at the driver’s seat. Although the accident circumstances still have to be clarified, the crash only raised the tone of a discussion on how Tesla tests its technologies. Waymo’s co-CEOs fear the company’s approach may slow down or even harm autonomous driving, according to their interview with Forbes.
Tekedra Mawakana and Dmitri Dolgov became co-CEOs at the autonomous vehicle company after John Krafcik stepped down on April 2. In her interview, Mawakana said that there would be no other “chance to build, develop, deploy, and introduce this technology again” because it is based on trust. If people lose their faith that it is safe and accurate, you may not have a second chance. According to the co-CEO, “you have to do it right the first time.”
While Waymo only uses its own trained personnel to deal with its autonomous prototypes, Tesla allows regular customers – such as the men who died in Texas – to test the software the company claims will allow its robotaxis to work as beta software.
While that could speed up the process and make it a lot less expensive for Tesla, it also places the responsibility on customers’ shoulders. Another "advantage" for the company is that it dodges the need for government test authorizations.
According to Mawakana, that's a huge risk that clients may confuse one system for the others. While Tesla states Autopilot and Full Self-Driving are Level 2 driving assistance systems – contrary to what their names suggest – Waymo is trying to deliver at least Level 4, which means no driver must be behind the wheel ready to take control whenever necessary. When a Level 4 system is ready, these cars will need no driver, and they will probably lack a steering wheel.
The risk of people confusing things and companies is real. Check this discussion between Warren Redlich and Taylor Ogan in which Redlich accuses Waymo of operating the car with some sort of remote control.
Redlich also said that the proof that Waymo’s tech does not work is that the company did not scale it. We’re not sure if Waymo’s co-CEO followed this controversy, but they have addressed it in their interview with Forbes. According to Mawakana, the company has to “build public trust,” “gain acceptance,” and “make sure the regulatory landscape is ripe and right for it.” When these three aspects are appropriate, and the technology is mature enough, “you can scale from there.” Waymo plans to license the technology, not to build it on its own.
Summing up, Waymo’s tech would work, but it would still be under development. The company will not spend its silver bullet for traffic safety allowing regular customers to test it instead of Waymo. Trust is the company’s most valuable asset, and it is not willing to waste it to speed up the process. Yet, it fears that will not be enough if other companies such as Tesla are willing to risk more than they should.