Driverless cars: who's in control?

Sunday 23 October, 12.00 - 13.00 , Frobisher Auditorium 1 Technology and ethics


Listen to this session at the bottom of this page.

The dream of driverless vehicles has existed since at least the 1930s when science fiction writer David H Keller wrote his story ‘The Living Machine’. The benefits seem clear; from allowing us the prospect of the more efficient use of travelling time to reducing the risks to car occupants, pedestrians and other road users. And the relevant technology has rapidly developed since the 2004 DARPA Grand Challenge, when none of the competitors completed the course and the furthest distance covered was a mere seven miles.  Over the past year or so, public road testing of fully automated systems have become common, and electric vehicle innovator Tesla has introduced a software update to some of its vehicle models allowing a semi-autonomous ‘autopilot’ mode. Driverless vehicles have moved beyond the ‘lab phase’ and are likely to appear on ‘a road near you’ soon. While automated vehicle pioneer Google has had its vehicles involved in only minor collisions, Tesla gained notoriety with the death of Joshua Brown driving a Model S, whose ‘autopilot’ controlled car failed to spot a large white truck. Many commentators who expressed concern over the maturity of the control technologies before that event might feel vindicated. Critics of fully automated systems question whether automated systems can ever make the split second moral judgements required in many situations.  Even if they could, can mere sensors provide the kind of sensory input that allows human to anticipate events?

More prosaically, traditional manufacturers like Ford, BMW, Volkswagen, Toyota and even GM may be in a better state than a decade ago, but profit margins are still generally low. And even if they can make driverless cars work economically, do they have the courage to deal with the legal minefields that a mass rollout of autonomous systems will involve? Do the ‘new kids on the block’ such as Google and Tesla in contrast have the attitude which could confidently take on such setbacks?

Is the impulse to develop automated vehicle systems more about squeezing more from the existing relatively dilapidated infrastructure, especially in urban areas, than developing a bigger and better infrastructure we need? Will driverless cars liberate us from gridlock and make driving safer, faster and more enjoyable? Or are they a technological dead end, economically impracticable and fundamentally unsafe?