The innovation race between car companies and tech giants like Google and Uber has seen expectations for driverless technology soar.
Proponents claim autonomous vehicles (AVs) can solve the problems we currently experience on the road: traffic will be safer, less congested and cleaner, there will be more car and ride-sharing, reduced labour costs in freight transport, and greater mobility and social participation among the disabled and elderly.
But will AVs achieve all this in 10 or even 25 years’ time?
There are good reasons to be sceptical. Spreading optimism about AVs makes sense if you are trying to generate support among regulators and investors. But the capabilities of new technology to have deep impacts on existing road transport systems are simply overestimated.
Being able to drive may not be as culturally significant as it once was but it is still important to the identity of many people, including many youngsters. And while driving is certainly not always enjoyable, the positive emotions and sensations it can generate are one reason why cars remain the dominant form of transport.
Some barriers to widespread uptake of AVs like insurance, culpability and liability in accidents, and the risk of hacking are well recognised. They are, however, framed as fixable instead of as wicked social controversies that might erupt once the public has gained first-hand experience of riding or sharing the road with AVs.
Industry and governments alike make strong assumptions about AVs’ desirability after market introduction. As long as AV developments revolve around vehicles and technology rather than people and everyday mobility, large-scale public resistance is a genuine risk.
Some of the promised benefits also seem too good to be true. If freight movements on the motorway or in cities become fully autonomous, drivers are unlikely to be eliminated in many – if not most – cases.
Drivers, after all, do so much more than drive: they manage loading and unloading operations, offer consumer services and intervene when things go awry. Replacing their roles would require far deeper changes to logistical systems than automation of driving.
Large-scale uptake of full autonomy may, for quite some time at least, be limited to places with simple and low traffic, for instance in ports or distribution centres, and to platooning of trucks on motorways.
Urban traffic is by far the most difficult to automate. Stop-and-go traffic and interactions with pedestrians, cyclists and other road users are challenging for all AVs but formidable for larger and heavier ones.
Given that conventional vehicle manufacturers have become key to AV developments, a large-scale shift from car ownership to car clubs and Uber-style ride sharing following automation looks increasingly doubtful. The car industry is venturing into “mobility services” like car sharing and smartphone apps for personalised mobility planning, but remains deeply locked into business models premised on individual ownership.
AVs may strengthen car- and ride-sharing in places where various forms of public transport and cycling are seamlessly interconnected. Yet, in the UK tightly integrated mobility systems are sparse, especially outside central London. Car-sharing is often more competitor than complement to public transport, which reflects a wider post-privatisation culture of transport service provision characterised by competition, short-term profitability and distrust of other players.
This makes the long-term growth potential of car-sharing uncertain. AVs may well re-entrench individual ownership and kill off “peak car” – the reversal since the 1990s of the historic growth in car use and ownership.
Moreover, it looks increasingly likely that AVs’ sensing and data-processing technologies will be integrated into vehicles themselves rather than the wider road infrastructure. The inevitable extra costs of AV technologies will therefore be shouldered by owners and users, and this will raise social justice issues.
The disabled and elderly, who might in theory benefit substantially from AVs, will be over-represented among groups for whom these vehicles will be unaffordable. The same holds for small operators in the freight and service sectors, including white van drivers.
At a time when many investments in public transport and cycling in UK cities contribute to gentrification and disproportionately favour the middle classes, as with HS2 and most bike-sharing schemes, AV developments risk further increasing transport’s role in enhancing social inequality.
AVs can reduce some of road transport’s problems, but only under specific conditions. Without careful, proactive and participatory planning, they may create and exacerbate more problems than they are expected to solve.
from Artificial intelligence (AI) | The Guardian http://ift.tt/2u3r2Ls