July 20, 2016

Autonomous Vehicles - Death by driverless car

As technology shifts to fully automated systems, liability is likely to be determined by the extent to which the driver is expected to remain in control of the vehicle

Car manufacturer Tesla is having a bad summer, with the forecast for a driverless car future looking a little less sunny. Tesla's semi-autonomous vehicle has had three crashes in as many months while in autopilot mode, one resulting in a fatality.

Tesla's autopilot software controls acceleration, steering and braking, allowing drivers to remove their hands from the steering wheel, subject to warnings for prolonged use. In these cases, it is alleged the drivers did not have their hands on the wheel at the time of the accident.

These incidents raise important liability questions concerning this emerging technology, which has the potential to disrupt and fundamentally change existing motor insurance models internationally.

At present, drivers of quasi-driverless vehicles are required to "stay in control" of the vehicle, even when in autopilot mode. A cursory glance on YouTube will show many drivers enjoying other activities for extended periods (including napping) when in autopilot. Vehicular warnings for steering wheel under-use are seemingly not given.

Regulators in the US are investigating the role driverless technology played in these accidents. Under current regulations, drivers will be at fault for an accident even when not driving the vehicle. However, the seemingly inadequate autopilot warning system, which should prevent extended use, may have contributed to the misplaced over-reliance in this technology.

These cases are indicative of the robust legal and insurance frameworks required, considering the liability position of both drivers and vehicles as the risks evolve. There has already been a plethora of cases brought in the US for driver assistance systems, indicating the law is able to respond appropriately.

The UK government's ongoing consultation on driverless cars notes the need for product liability insurance to cover injuries to the innocent automated vehicle driver as well as passengers and third parties. A two-tier insurance regime covering drivers and driverless cars will be required until consumer confidence is secured. Presently 50% of consumers do not trust driverless cars.

As technology shifts from driver assistance to fully automated systems, liability is likely to be determined by the extent to which the driver is expected to remain in control of the vehicle. However, the liability position during the "handover" period between driver and autonomous modes remains unclear. Liability could be shared between the driver, manufacturer, software provider, software engineer and vehicle data provider in such circumstances.

A more complex network of cover will create a more complex claims resolution process. Motor insurers may continue to settle claims and seek an indemnity from the manufacturer or others. Lawyers will need to pursue claims against many different entities, resulting in increased legal costs and protracted litigation.

Coverage disputes in the motor market may become commonplace. Existing policy wordings will need revisiting to ensure policies respond to the developing risks. The first driverless policy was launched recently, covering loss for hacking, system and software failures. Although it is not expected to be employed in the short term, the policy is being used as a benchmark by the industry to prompt discussion concerning the development of legislation.

Indeed, driverless vehicles represent an opportunity for insurers on an international scale. The developing technology provides the prospect of harmonising motor insurance regulation on a global basis. This has the potential to afford insurers the chance to enter previously untapped markets where motor insurance is still not compulsory.