ADAS already proving the UBI case

ADAS features are becoming more commonplace on new vehicles rolling off production lines and while the systems function to help drivers navigate more safely, it’s unclear just how much visibility insurance carriers have on the performance of these next generation enhancements.

“Most insurers are happy to give discounts for new features that are proven to reduce their costs but when systems are new such as many ADAS options, there isn’t the claims experience available at the time when the systems are introduced,” notes Andy Goldby, chief product officer for telematics data management firm The Floow.

All this will change as carmaker telematics data becomes available and insurance companies can readily identify which systems were fitted, active and activated at any given time and this will then enable insurers to rate them far more accurately.“Most insurers have done limited analysis of the loss cost savings from ADAS equipment. These savings will eventually be built into vehicle symbol rating through the standard processes, once sufficient experience emerges,” says Robin Harbage, director, Willis Towers Watson.

Since the ADAS systems are in a nascent stage and more data must be collected, Harbage notes: “In the meantime, HLDI has provided significant guidance in the expected loss cost savings from ADAS equipment.”

Sweden’s If P&C Insurance and Volvo jointly produced one of the first deep dives into how some ADAS features are functioning in the real world, evaluating rear-end collision avoidance technologies. The study looked at cars with forward collision warning and brake support combined with adaptive cruise control (CWB + ACC). It used data from insurance claims and showed that vehicles equipped with CWB+ACC saw 38% fewer rear-end crashes with frontal impacts.

The research also showed that the latest technology at that time was even more effective at avoiding crashes than prior models. It found a collision avoidance rate of 45% for the third-generation features. Volvo executives point to this collaboration and others, noting a long tradition of the manufacturer working with insurance companies. Eva Lahti, product attribute senior manager, driver assistance and safetyat Volvo says the cooperation is expected to increase going forward “for the task of measuring the performance of the ADAS and its ability to minimise risks.”

Lahti points to an insurance study that covered more than 100,000 insured vehicle years that found cars equipped with the carmaker’s City Safety featured were involved in 27% fewer accidents and subsequent insurance claims. Clearly greater cooperation will enhance these studies. “Since these features have not been available on the market for a long time and in large volumes yet, many technologies were not possible to evaluate,” explains Lahti. However, she adds that it will be forthcoming: “After the summer holidays, a report on the LDW effectiveness will be available. Also, blind spot Information systems and pedestrian detection systems are waiting for their real-world follow-up.”

Clearly the more data the better but what can be confounding to researchers and insurance companies is differentiating how better, and safer, driving is achieved. At this stage there’s not enough data to break down whether the driver or the ADAS is have greater effect on the car’s manoeuvring and safety. “Most UBI products are focused on the vehicle operation,” says Harbage. “They do not differentiate between vehicle safety as a result of good driving and safety resulting from ADAS equipment.”

While it may take a while for insurance companies to sort that out, Harbage notes “if an ADAS feature improves the safe handling of the vehicle, it will be directly measured in the current policy period and applied in a discount during the next renewal.” He says insurers can work with carmakers to decode the VINs and analyse ADAS features. This can help streamline the assessment process for connected and autonomous vehicles, especially when you have models that are fitted with ADAS as an option. Yet, even with this ability, it currently doesn’t provide 100% clarity into whether the features are functioning or manually switched off in some cases.

Goldby says this presents a unique challenge: “It is much harder to identify whether it is actually present on a particular vehicle.  This is further compounded by the fact than many ADAS systems can be disabled by the user – think turning off ESP – and, therefore, it is not currently an easy job to link claims to the experience of new ADAS systems since you have to be certain that they were active at the time.”

Conditional automation in features such as adaptive cruise control, assisted braking, and lane assistance will lead to a dependence on UBI data for the most effective pricing. Goldby says: “Over time this will become important especially if the OEMs start to accept liability when the car is in autonomous mode.  In these cases UBI data will be required form the vehicle so that we can identify when the human is in control and when the vehicle is driving and apportion the scores and liability correctly.”

He added that work is being completed to provide data accessed from the CAN bus to create an ADAS-based ‘vehicle’ behaviour score which would calculate the total risk of the car and operator. “However, it does require good quality UBI data sourced directly from the vehicle to provide certainty to who was in control for every section of the trip.”

As for accessing data in the small but growing UBI consumer market, some insurance companies are collecting real-time data and Harbage says firms are working to fill that void. “The existence of large UBI databases … can directly measure ADAS effectiveness of loss cost reduction.”

[Ins.Gray.2017.05.12]