It seems we've come a long way in the last few years. The dorky looking Google cars really haven't been around for that long but now they seem, at least to me, "normal" whenever I see one on the TV. Something that is a little new is the fact that Perth is set to be a pioneer in driverless ride-sharing tech in a trial set to start next year according to this article.
I'll need to reuse the word "dorky" here to describe the look of the cars in question and pictured here courtesy of abc.net.au. They're the product of a French company called NAVYA (which I'd not heard of prior to this article, another telling observation about this technology space). According to the article, the Western Australian Government has partnered with NAVYA to bring several driverless cars to the state for testing next year.
abc.net.au
Operating in much the same way as ride-sharing services like Uber, customers will be able to order a ride using a smart-phone app.A car will then arrive to take them wherever they need to go — only without a driver.
I'm really not sure how I feel about this. Certainly progress has it's risks and costs and working in technology, I have a fairly firm idea that we'll one day see really usable driverless tech but a public trial at this stage makes me cringe a little.
When operating automatically, the vehicles employ multi-sensor technology including four radars, six cameras and two links to global navigation satellite systems. A black glass or plastic screen on a box bolted into the car's roof.The features give the vehicles 3D perception that allows them to map the environment, detect obstacles and respond accordingly. RAC WA's CEO Terry Agnew said it was hoped driverless technology could one day eliminate the 90 per cent of road crashes that were caused by human error.
He said the technology would tie in nicely with Perth's existing transport systems, and suggested the trial could put the city at the forefront of the "driverless revolution".
"The driverless passenger vehicles will largely be shorter journeys, they'll be inner-city and very often they might be simply connecting people up to public transport," he said
We are likely all familiar with the Tesla incidents involving autopilot technology and keep in mind that those have occurred with drivers behind the wheel, presumably ready to observe and intervene in any dangerous situation (although I agree that in reality this may be of limited impact). I also happen to drive a vehicle with some semi-autonomous features and I can personally attest to the fact that even simple radar cruise control can be tricked by fairly common scenarios.
Take the example of two lanes that go around a shallow bend. One lane (the inner) may be moving at speed and the second (outer) land may be a turn lane with stationary traffic. Such a setup passes me by on my daily commute. Radar cruise in use in the inner through lane will detect stationary traffic in the outer lane ahead of it as the car sweeps around the bend and start braking heavily thinking that it is a collision scenario. It is only once the direction of the vehicle sweeps past the cars in the outer lane that things remedy themselves. A simple and common scenario involving only brake application. Throw steering into the mix and the inability to intervene and it does really make me wonder.
Unless we're talking about putting around in the suburbs (which is the proposal here) I have my doubts about whether we're at the point of public ride-sharing trials. What do you think? Would you use one? Are we ahead of ourselves or is the risk the cost of progress? Should we be trialing these vehicles where there is less risk of pedestrian injury? It's a really interesting topic and I look forward to the discussion.