Who owns the data?

As nations globally instigated lockdown measures in Spring 2020 to control the spread of COVID-19, private cars largely went unused for weeks on end. The annual motor insurance premium calculated on a range of factors such as where the policyholder lives, their age, occupation and engine size could not factor for this sudden and prolonged change in risk. Insurance providers offering telematics or usage-based insurance products had insight the rest of the market could only guess at with reports of traffic volumes falling between 70%-85%, as cited in analysis from LexisNexis Risk Solutions and leading Telematics UK broker Autosaint of March to June 2020.  

The pandemic has highlighted the value of static and dynamic vehicle data in understanding risk, delivering a fair and accurate premium and giving motorists more flexibility and control over their insurance costs. Rather than rely on estimations and proxies for risk, it could tell an insurer the car’s mileage reading, valuable insight during and post-pandemic, the car’s precise location, how it is driven, and the Advanced Driver Assistance Systems (ADAS) features on-board and activated to enable the development of more personalised, more engaging insurance cover. This is not visionary: the insurance and car manufacturing markets were coming together to make this concept a reality, well before the emergence of COVID-19. 

Graph: Lexus Nexus

The big question has been how to bring these two major industries together so that insurance providers can access connected car data in a way that adheres to data privacy and security regulations as well as making sure that the data is useable and meaningful for insurance.

How do you ensure a consumer driving a car is given the option to share their driving data for insurance and this will be understood in the same way as their neighbour driving a completely different car? When you think about the volume of car makes and models, the number of motor insurance providers, and the quantity of motorists buying insurance, it becomes a big ‘many to many’ problem.

The solution lies in a data platform strategy that takes all the strands of vehicle-centric data, telematics data from any device or app, data from the connected car and vehicle build data then puts it into an environment where it can be standardised, contextualised, normalised and scored in a fully compliant manner for use across a variety of vehicle makes or models.

This is one Big Data project that could have a profound impact on the future of mobility.

The starting point goes back to the investment car makers are putting in vehicle autonomy with Advanced Driver Assistance Systems (ADAS) increasingly available in vehicles and constantly being enhanced. SMMT research shows 8 in 10 new cars in the UK have driver assistance systems and over half have adaptive cruise control. In the US, internal analysis by LexisNexis Risk Solutions shows 76% of 2019 models had at least one core ADAS feature, which is up significantly compared to 18% in 2014.

To date, it has been a challenge for insurers to identify exactly what ADAS features a specific vehicle is equipped with when writing a motor insurance policy. This is because each car manufacturer has created its own unique terminology, definitions, and naming structures – sometimes releasing multiple features within the same model year. In addition, many items are chosen as optional extras when a vehicle is purchased new. 

To address this challenge, data scientists at LexisNexis Risk Solutions have developed an ADAS classification system using machine learning to scan millions of lines of car manufacturer vehicle data to logically sequence and classify vehicle safety features and component’s intended operation or purpose. This classification system provides the foundation for LexisNexis® Vehicle Build.

Access to vehicle safety data will help insurance providers factor for their presence throughout the customer journey – in pricing, mid-term adjustments and renewals – and establish the differences in risk profile associated with the vehicles that have these safety features. It will also enable them to understand how specific safety features behave. For example, if a feature will provide an alert or warning to the vehicle’s driver when a potential danger or hazard is detected or if it can take active control of breaks and steering. It will allow insurers to understand the purpose of features.

At the same time, car manufacturers will be able to see what ADAS features have the most impact on their customer’s safety to help drive further enhancements.

As carmakers and insurers get a better understanding of the vehicle, how it’s driven, its performance and how semi-autonomous features work in the real world, consumers will also learn how data from and about the car can work for them. This will support market adoption, drive innovation, and help create more benefits in terms of safety and total cost of ownership.

Image: smashHIT Project