COMMENT: Uber’s mess won’t derail autonomy, but industry must rein in the hype

You’d be forgiven for thinking full autonomy on public roads was just around the corner. The self-driving Uber crash in Tempe, Arizona, proves conclusively otherwise. By Xavier Boucherat

Uber has proven to be one of the most disruptive and polarising forces in the mobility industry. The company has never been far from scandal, with driver disagreements, CEO public meltdowns, technology theft allegations and stories of a toxic work culture all making the headlines.

Yet through it all, Uber has retained an ‘innovator’ status, lately thanks to its efforts to bring self-driving vehicles to the road. Test cars have become a common sight in states such as Arizona, and stunts like the world’s first automated delivery of beer in 2016 have also sought to depict the company as a leader in the field.

It begs the question, what stage is the industry really at with autonomous vehicles? Indeed, what is an autonomous vehicle? Does the example of Rafaela Vasquez show that the human-machine model proposed by Level 3 systems is untenable?

But following the events of March 2018, when 49-year old Elaine Herzberg was struck and killed by a company test vehicle, it is all too clear the technology was not ready for public roads. It was stark proof that Uber lags behind rivals such as Waymo and Cruise, and whilst preliminary investigations have yet to establish probable cause, the technical shortcomings are evident – the vehicle failed to recognise emergency braking was required until far too late, and furthermore failed to detect that safety driver Rafaela Vasquez was distracted, and bring her back into the loop.

There can be no doubt that Herzberg’s decision to not use a road crossing on a dark stretch of road created a challenging situation for the vehicle, but dashcam footage released by the Tempe police department shows her coming into view some seconds before impact – arguably enough time for an attentive driver to brake and take at least some speed out of the vehicle. What’s more, it is thought that an unmodified Volvo XC90, with its advanced driver assistance systems (ADAS) active, could have performed better in the scenario.

It begs the question, what stage is the industry really at with autonomous vehicles? Indeed, what is an autonomous vehicle? Does the example of Rafaela Vasquez show that the human-machine model proposed by Level 3 systems is untenable? How must current sensor and monitoring technology be improved to be sure the same mistake doesn’t happen again? And how best does the industry conduct public testing moving forward?

In recent years, however, the competition to bring a self-driving vehicle to the road has become notably fiercer. The industry must ensure such pressure doesn’t prompt the premature introduction of the technology

If major automakers weren’t considering these issues before, they certainly are now. The societal benefits of autonomous driving, in terms of safety, efficiency and accessibility, remain large, and no truly significant setbacks have come about as a result of the Tempe crash.

In recent years, however, the competition to bring a self-driving vehicle to the road has become notably fiercer. The industry must ensure such pressure doesn’t prompt the premature introduction of the technology, which could only further complicate the challenge of public acceptance. Nor must it hype technology beyond its capabilities – consumer misunderstandings have already proven fatal in the past. To learn more about how the industry is learning from the Tempe incident, and what other concerns it might have prompted, download Automotive World’s Special report: Implications of the Phoenix Uber crash.

Close
Close