Assisted and Automated driving poses immediate challenges for international regulators

The last twenty years has seen a dramatic improvement in the safety features protecting drivers involved in a car crash, with seatbelts, multiple airbags and side protection beams on every new model. But we’re now moving beyond the age of passive safety (protection in the event of a crash) into active features such as automated braking and lane control systems that assist the driver to help avoid the crash in the first place. The technology journey continues as these functionalities are now being combined to allow the vehicles to drive themselves (automated driving) to the point where crashes could be prevented altogether. However, if these benefits are to be realised, we must not lose focus on the driver’s understanding of what these various systems are, and are not, and what the drivers obligations might be. Are they in control or the vehicle? In the split seconds before a potential crash, will they be available to take back control of the vehicle quick enough if the system can’t cope?

Vehicle manufacturers are actively preparing to roll-out highly advanced automated driving features within the next few years – the first wave from 2019 will provide the driver with more assistance and some limited automation.  From 2021 onwards, we’ll see technology capable of taking over entirely from the driver on certain roads. By 2025 the car may be able to drive door to door without driver intervention.

Thatcham Research has been using insurance industry data and road safety statistics to develop a ‘Claims of the Future’ model, which we are developing to help insurers prepare for this technology. All our projections show a positive impact for this technology, with an overall crash reduction of circa 20% over the next 10-15 years.

However, before we reach full automation, there is likely to be a transition phase where drivers are required to perform less and less of the driving task, but are still obliged to take back control. This may seem like a logical approach at first, as it seemingly allows drivers to get used to increased autonomy gradually. However, it may create new risks that the drivers and insurers are not prepared for. When cars come with such a high degree of “Assisted Driving” functionality (but are not fully automated) they appear to be ‘driving themselves’. But the driver may not be aware that they still play a crucial role in monitoring this automation, and are still required to take back control in the event of a sensor failure or when at the end of an automated domain. This misunderstanding may indeed end up creating more crashes, and many of these could be catastrophic high speed crashes on motorways.

Regulations governing the construction and use of vehicles are set internationally (within the UNECE’s working party 29) and the process of changing these regulations is complex. Arguably, this process was never intended to address the kind of rapidly evolving technological challenges it now faces. Most of the key decisions are focused on assisted and automated steering systems, but it is clear that automated driving will affect every aspect of a vehicle’s functionality including braking and signalling.  We believe that a dedicated work stream that adopts a more streamlined view of automated technology is now needed if the benefits of these systems are to be realised quickly, but still within a regulated environment.

Thatcham Research has been working with insurers to develop a set of proposals for how this “Assisted Driving” functionality must operate, to prevent drivers either intentionally or unintentionally misusing this technology. Our proposals for minimum standards for any “Assisted Driving” vehicle that receives type approval from international regulators include geo-fencing, so it can only be used in certain locations, ensuring automatic breaking and collision avoidance systems are always active, monitoring when the driver is attentive and warning them when they’re not, and allowing the vehicle to stop safely at the side of the road in cases where the driver fails to respond to a takeover request. Later “Automated Driving” systems should, by their very nature, have these capabilities and not require the driver to intervene to remain safe, but the differentiation in these capabilities must remain clear.

There need to be clear standards for what drivers are expected to do when using this technology. If regulators do not address this, there is a risk that certain manufacturers will feel the process is holding them back and seek ways to ‘get around’ the regulatory regime as some have appeared to do already. Much of this technology carries exciting safety benefits, and the insurance industry wants to encourage these systems onto the roads, but it is vital that regulation keeps pace so that the rollout of automated technology can be managed safely.

Matthew Avery will be giving a keynote presentation on Thatcham Research’s work to model the impact of automated driving and influence the regulatory framework for vehicle technology at the ABI’s “Changing Gear – Adapting to Autonomous Vehicles” event on 6^th November. To book your tickets, click here.