The increasing level of autonomy of vehicle systems such as steering and speed control and greater connectivity between vehicles is already bringing significant new challenges across many areas of the automotive industry.
The accurate assessment of likely risks in a populated road environment is not only essential to keep insurance premiums competitive for the consumer, but to ensure that appropriate long term planning decisions are taken by legislators, vehicle manufacturers and repairers. So it’s no surprise that the insurance industry is at the forefront of actions to better understand future traffic scenarios.
The rate at which autonomous technology arrives on the highway will be influenced by the legislative environment in each individual market. For example, the legal framework in the UK is quite favourable because it is based upon the precept – common law – that something is ‘legal’ unless specifically prohibited. Many other markets are governed by the opposite argument and require specific legislation in order to enable the application and use of the new technology. By producing reliable simulation data to demonstrate the likely impact of introducing greater levels of autonomous functionality, we can help policy makers to prioritise appropriate legislation that will support, rather than obstruct, the efforts of the automotive industry.
It has been suggested that the technical challenges of autonomous vehicles, interconnected and communicating with each other, are easier to address when the entire vehicle population is autonomous, than during the transition when they must co-exist with manual vehicles. This may be so but, from a risk perspective, we already share the highway with drivers of mixed abilities who often behave unpredictably. Every autonomous vehicle that enters service displaces a manually driven one, replacing a human driver who may be distracted, tired or in a hurry with a machine that will always prioritise safety when making decisions. This means that even during the transition phase when the vehicle population is mixed, autonomous vehicles are likely to contribute to improved road safety by incrementally reducing risk. There are also as yet unknown risks which may become apparent when the driven fleet interacts with the driverless. It is impossible to predict or model these, the best we can do is to reduce the estimate of the potential road safety benefits to factor for these. However, given the relatively poor competence of human drivers, we should not be surprised if these do not materialise!
At Thatcham Research – the motor insurers’ vehicle research centre in the UK, usually referred to simply as ‘Thatcham’ – we have begun a long- term modelling project to examine how the imminent arrival of digital technologies will affect the types of accidents that occur, their frequency and severity. The initial indications from this ‘Claim of the Future’ study are highly encouraging, suggesting a positive effect on safety, increased personal mobility and even a potential increase in national GDP by reducing the direct and indirect costs associated with vehicle collisions. Interestingly, this is supported by a recent report from KPMG which identified that the overall economic and social benefit of connected and autonomous vehicles in the UK alone could be in the region of £51bn (US$80bn) per year by 2030.
That’s only 15 years hence, and reflects the growing realisation that autonomous control is no longer a distant prospect. Lane keeping and intelligent speed control are already with us on the highway, while in the city assisted braking and guided parking are becoming common options: BMW recently demonstrated its Remote Valet Parking Assistant that enables parking while the driver is outside the vehicle, using a smart phone app; other vehicle manufacturers are championing similar technologies.
The next steps are to begin linking these relatively safe and easily managed functions into a wider self- driving capability. Insurers, who charge the individual consumer a few hundred dollars to cover a risk that could run into millions, traditionally manage their financial exposure by consideration of the vehicle type and driver profile when calculating premiums. The new technologies are many generations on from simple warning alerts and have the clear potential to reduce the threat posed by inattentive, incompetent or aggressive drivers. Unsurprisingly, they could also help neutralise the dangers from inappropriate use of excessive speed in high performance vehicles.
The effect on the repair industry is likely to reflect fewer collisions, thanks to better collision avoidance, but an increase in the complexity of the repair process. On a highly digitised vehicle, this will entail management of a number of interacting electronic systems and thorough re-validation before returning a repaired vehicle to the customer. While the physical volume of repair work is likely to be much reduced, the level of skill and facilities required to ensure a satisfactory repair will be much wider, and the total value in the repair chain may not reduce proportionately with the reduction in crash frequency which is promised by the autonomous vehicle technologies.
Another issue that is frequently discussed concerns legal liability following a collision involving an autonomous vehicle. We already have situations where the manufacturer of a vehicle or a system which malfunctions can be held legally responsible, so what’s the difference? The high levels of connectivity associated with autonomous operation mean we will have unprecedented amounts of data available, making it easier than ever before to analyse events leading up to a collision and apportion appropriate liability – and some of the current road testing strategies include making provision of telematics data a requirement for highly and fully autonomous vehicles.
Of course, the unprecedented quantity of data will require equally unprecedented system capacity to communicate it all, and this could certainly limit the rate at which autonomous vehicles can be introduced. At present, 3G network coverage is variable; in the future we are going to need 4G or 5G to provide the necessary level of connectivity. The telecoms providers need reliable forecasts of likely future demand in order to support timely investment decisions; otherwise, the mobility of an entire region could be affected. Governments must urgently review this in their future considerations of the roll- out of these technologies.
In predicting demand, the information generated by organisations like Thatcham helps provide useful supporting data, but ultimately the reaction of the consumer will always be tough to predict; as the saying often attributed to Henry Ford goes, “If I’d asked what my customers wanted, they’d have told me ‘a faster horse’”. We in the automotive industry may well be surprised at the rate of growth, because this new dimension in personal mobility has appeal for all age groups. Older customers with diminishing competence can rediscover the freedom of independent travel without the stress; younger people who may opt to forgo driving if it curtails their connectivity with others will no longer have to make the choice between one or the other.
On the other hand, consumer confidence in new technology can be fragile. A recent poll found that many adults who expressed a willingness to travel in an autonomous vehicle were less enthusiastic when asked if they would let their children travel in one. Faced with such conflicting indicators, the greater the amount of reliable information we can share, the better will be the quality of our long- term planning. The revolution is here – let’s be ready for it.
Andrew Miller
