Failing or poorly performing technology is nothing new. Here, Antony Savvas considers the fallout for fleet management as a result of the recent autonomous Uber vehicle fatality in the US. The technology industry and the hype that surrounds it is probably not matched by any other industry.
Very often both business and consumer users get sucked into the need to have a new technology right now, even though the signs are there for everyone to see that it is not really ready.
Go back to the first broadband services becoming available and after that the first smartphones being launched, and the marketing and advertising being gratefully swallowed by the media.
The first DSL broadband services did not reliably allow you to download movies onto your TV or desktop on demand, in fact they would have struggled with short YouTube clips. And the first data phones did not enable you to easily “surf the mobile internet”. This was despite many signing up for these expensive services. But while people were left disappointed, nobody was hurt.
Thankfully, it doesn’t matter how much money you have, you can’t just ring up and order a driverless vehicle to jump into. There really is a comprehensive development process that has to be followed through first, that has to work within national and international parameters with safety the main obvious consideration.
Autonomous vehicles (AVs) may be making progress within that process, but if that progress is to become rapid, those vehicles have to master three main tasks: understand their environment, predict the actions of those (or things) around them and accordingly respond.
Arguably, those first two tasks are the most difficult to complete as far as technology development is concerned, even as artificial intelligence is apparently spreading around us like wildfire (according to the hype, again).
For commercial fleet management, the tragic fatality in Tempe, Arizona this month, involving an Uber autonomous vehicle and a female pedestrian, may have been seen by some as a defining or even ruinous moment for the commercial deployment of such services.
The car and truck firms, a multitude of start-ups, and not least Uber and Google‘s Waymo, are waiting to see if there is any major effect on their developing driverless business as a result of the accident.
Waymo for one though is adamant – and some might say brave – in declaring that its technology could have prevented the Arizona fatality.
At a gathering of the National Automobile Dealers Association in Las Vegas last week, Waymo CEO John Krafcik told delegates that Waymo was already well advanced towards its goal of decreasing motor vehicle fatalities.
And to emphasise his point, referring to Tempe, Arizona, Krafcik said: “We have a lot of confidence that our technology would be able to handle a situation like that.”
That “situation”, under investigation by police and federal safety regulators, saw an Uber test vehicle driving in autonomous mode – with a back-up driver on board – hit the female pedestrian at night, as she was walking across a four-lane road with her bike.
While we must look further into the detail, the latest facts emanating from the incident are disturbing. Uber has just been banned from carrying out further tests in Arizona until the safety investigation is completed, with state governor Doug Ducey referencing a released a video showing the Uber back-up driver not watching the road before the collision.
So, is this a failure of technology or a failure of humans to use technology properly?
Waymo has been working on its AV system since 2009, with 5 million miles of testing on public roads having been completed by the firm, it says. Significantly, the Waymo car AV system is also basically the same as the one which is deployed in its self-driving trucks which started commercial operations this month.
One can only hope that Waymo and others can truly demonstrate how the advantages of a very hyped technology can firmly outweigh, not the chance of disappointment, but the chance of extreme danger and death.
The author of this blog is Antony Savvas, a freelance technology writer