Do self-driving cars and humans mix?

The Google Self-Driving Car Project holds a lot of promise for the future of transportation. Autonomous vehicles could increase safety on the roads, and would potentially enable the visually impaired and the elderly to remain independent, and mobile beyond their years of safe driving.

To date Google cars have been involved in sixteen reported minor accidents. Twelve were reported in the May 2015 Google Self-Driving Car Project Report. A further two accidents were noted in the June report and one more each in the July and August reports.

According to Google, in all cases the vehicle itself was not at fault because the cars were either being manually driven at the time, or the driver of another vehicle was at fault.

It would appear the biggest enemy of Google cars are … humans. Distracted drivers to be more exact, who have a tendency to rear end the self-driving cars.

There has been a more amusing report of the Google car’s first encounter with a roundabout, where the car decided the safest thing to do was to keep going around. Chris Urmson, the man behind Google’s self-driving cars recounted the event: ‘There were a couple of engineers in the car who were giddy going round and round. It felt very Chevy Chase-esque.’

When it comes to the roundabout incident, I empathise with the poor Google car. I still remember my first encounter with a roundabout, and it didn’t go much better than the Google car’s first attempt. If it wasn’t for a break in the traffic I would probably still be going around …

The monthly Google Self-Driving Car Project reports disclose a number of other amusing experiences the Google car had in recent months, including the case of the wheelchair-bound grandmother chasing ducks in the middle of the street, which was handled by the Google car well.

Urmson, speaking at the Automated Vehicle Symposium in Ypsilanti, Michigan earlier this year, noted Google is now focusing on more complex and unpredictable driving situations that occur in urban street settings, rather than freeway driving conditions.

The project is of high significance due to its potential to reduce accidents, road fatalities, insurance claims and even traffic jams, provided there is a statistically large enough number of autonomous cars on our roads to achieve those outcomes. In that respect, achieving the forecast benefits of autonomous cars is a little bit like the relationship between immunisation and herd immunity – a minimum percentage of overall immunisation is required for the benefits of herd immunity to take effect.

As usual with emerging technologies, autonomous vehicles are likely to have a significant flow on effect on the economy. They will disrupt a number of industries, such as car repair services, the insurance industry, and the transportation sector from taxis to the trucking business, even public transport. The adjustment will likely be very painful because the technology has the potential to make millions of jobs redundant around the world.

Car insurance is a global, multibillion dollar industry. From an insurance perspective autonomous cars will largely do away with driver error, which could represent a serious threat to a substantial income-stream for the industry.

However, such cars will present their own unique vulnerabilities, from potential algorithm and software errors to hacking and privacy. Zurich produced a report recently titled ‘Smart cars and connected vehicles‘ which highlighted numerous potential issues identified by insurers.

With autonomous vehicles pre-programmed to obey speed limits, red lights and other traffic rules, traffic law enforcement, and income streams from traffic violations, would also be significantly affected. While this would free up police resources to deal with more serious matters, it will still cause significant disruption to police numbers and government revenue.

No doubt this latest leap in technology will frighten many because it’s … new technology.

Others will lament the potential long-term threat to the pleasure of driving. Currently, the car industry builds many of its campaigns around the ‘thrill’ of driving itself, from BMW’s ‘Ultimate Driving Machine’ tagline to Volkswagen’s ‘Drivers Wanted’ campaign.

Jay Samit of TechCrunch recently predicted driving would be illegal by 2030. While that timeline may be ambitious, there is certainly a kernel of truth to the prediction.

Samit’s views are supported by no less than Elon Musk, who also suggested speaking at NVIDIA’s GPU Technology Conference in San Jose, California, that ‘[i]n the distant future, [legislators] may outlaw driven cars because they’re too dangerous.’

It will be interesting to see how traditional car manufacturers and driving enthusiasts will respond to the threat presented by self-driving cars, and how insurers and legislatures will manage the implications of the technology once it’s ready for mass-market distribution. Arguably, the likely trend for legal and regulatory liability will be a shift towards the manufactures of self-driving cars, and away from consumers.