Driving the Next in motor risk

Recent algorithms are increasingly used for prediction: we are moving towards an algorithmic assessment of driver and vehicle risk.

Algorithmic predictions are:

• Individualized: individualized risk forecasting allows insurance companies to propose personalized offers, messages, pricing and recommendations. Which is the individual: the driver or the vehicle? Who owns the data?
• Performative: instead of compensating for the damage once it has occurred, the task of insurance would be to prevent it from happening. Driver coaching features: control or motivation?
• Opaque: we can take advantage of predictions we do not understand. How can policyholders be presented with a decision based on incomprehensible procedures?

• Carlos Pereira, director of the areas of Organization, Reinsurance and Innovation at CA Seguros, presented CA Best Drive, the telematics insurance product powered by Swiss Re’s Coloride.
• During the session, he presented the main features of the app. A quick overview can be seen here: Coloride: the app based telematics solution by Swiss Re and Movingdots on Vimeo
• An insurer like CA Seguros needs to have such an app to be able to deploy telematics in Portugal and support their clients in improving the quality of drive.

• Some of the most relevant behavioural factors that affect accident risk are drivers' overconfidence and overoptimism, leading to: smartphone usage, speeding and critical maneuvers. For example, writing a message on the phone can more than double the risk of accident and injury.
• Insights about these risk factors were used to behaviorally inform Swiss Re's telematics solution Coloride. Some of the behavioural economics principles contained in Coloride are anchoring/ loss aversion, immediate gratification, social norms, framing/ loss aversion/ nudging, continuous engagement and reinforcement, altruism/reciprocity.
• The impact of behavioral interventions to mitigate driver risk looks positive so far. After 30 trips we see 5% improvement of driving performance, 6% improvement for speeding, 4% improvement for distraction and 14% improvement for maneuvers.

• Predict is an Insurtech start-up founded in 2019. They provide a scoring system to help insurers accurately price commercial fleet risk. Predict uses tachographs and a highly technological process to get drivers’ driving style and trucks/logistics data.
• Benefits are shared: insurers have visibility over the fleet risk and can influence the future loss record; the fleet company also has visibility over the risk of their fleets and can implement remediation measures and training.
• The solution is live in Italy since 2021, with the goal to scale it first in EU and then globally.

• MOTER was founded in April 2021 as a subsidiary of Aioi Nissay Dowa Insurance Company, Ltd. (part of the MS&AD Insurance Group). MOTER is an edge-computing based data science platform for connected car data.
• There's a multi-billion dollar gap between auto makers and insurance carriers to effectively leverage connected car data, due to 1. High data costs & privacy risks 2. Massive regulatory hurdles 3. Complex & customized data requirements.
• This is where MOTER comes in, processing next-gen in-car data via edge computing to provide insights on driving behavior, environmental context, distracted driver and ADAS.

• Telematics (TMX) has demonstrated concrete impact to insurers:Productivity: TMX represents 30% of new business for top US auto insurers
  • Profitability: the “behavioral change” global best practice has measured on average 15% improvement in driving behaviors within 30 days (even doubled in the long term)
  • Persistency: 5-15 percentage points higher retention rates
  • Proximity: more interaction between insurers and clients
• Market maturity is very different from market to market. Mobile-based TMX has been the protagonist of the market evolution in the US, but it is not leveraged adequately in European markets yet.
• TMX is a business capability (not a product) : telematics doesn’t create value on its own, data needs to be actively used to generate value

• Telematics (behavioural) data and coaching strategies allow insurers to mitigate information asymmetry and gain insight into individual risk profiles, alleviating adverse selection.
• Telematics can encourage people to drive more safely. But is the change in driving behaviour the result of control or motivation?
• Insurers can try to control drivers' behavior with a combination of feedback and financial incentives (e.g., discounts or vouchers). But it is difficult to say with certainty whether the change in behavior was the result of coaching, the desire to save money or simply chance.
• Conclusion: coaching programs are one of the most promising implementations, but they
• require further research from a sociological perspective.

• Zego started in 2016 by focusing on delivery drivers; they are now the leader in flexible commercial motor insurance.
• They have developed the Zego Sense app: it provides insights and risk scoring thanks to telematics data.
• Zego has also developed a Fleet Portal for Fleet managers, helping them to control the driver risks and influence drivers to be safer.
• Zego's 2040 vision: thinking about how the new autonomous vehicles and robo taxis will impact commercial lines and the overall mobility.