How to build an IoT&Data Platform for transforming the insurance industry with connected cars Enrico Mancin Watson IoT Europe IBM Turin, Italy enrico.mancin@it.ibm.com Copyright © held by the author Abstract— The Systems Engineering approach and practices insured cars to track vehicles and be immediately informed have been key success factors for developing Telematic Hub for about car accident events. Soon, Insurance Companies have Insurance Industry as a real solution for production helping to been collecting a huge amount of data with the imperative need meet customer specific requirements since the first provided to analyze raw data for building and improving over the time input. Telematic Hub aims to leverage data coming from possible their own algorithms. existing motor solutions and to extend/expand telematic capabilities in order to easily respond and adapt to insurance industry business model changes. This innovative solution offers II. OBJECTIVES full support for three logical layers: Connected Devices, IoT In order to achieve their aim of becoming more effective at- Analytics and Big Data. risk selection and more efficient at handling claims, Insurance Companies needed more sophisticated techniques to analyze Systems Engineering adopted practices will be discussed for data. Through the black boxes installed in the vehicles and the developing the Telematic Hub solution going from concept to IBM IoT and Analytics Platform, an Insurance company can production and operation. collect data like: • The GPS position of the car; I. INTRODUCTION In Insurance Industry, Telematics is a “game changer” for • Driver behavior during regular trips (driver profiling); car insurance business: it adds value along the whole value • Specific events (e.g. to understand the dynamics of chain from customer acquisition through to pricing and claims car accidents); management. This helps Insurance companies to understand the need to revise product strategy, implementing large IoT These elements were being measured for two key reasons: programs in order to recover Revenue with new telematics- (1) to limit the number of false claims and (2) to understand the based product and to reduce risk costs for Claims and Crashes. dynamics of accidents (in attempt to predict and prevent). Insurers in Europe have experienced fraud on an increasing The adoption of SE Lifecyle and SE practices allowed the scale in their claims processing over recent years. Insurance development team to meet customer objectives, short time Europe, the European (re)insurance federation, estimates that delivery requirement and budget fully achieving all required the total from all cases of fraud – both detected and undetected functionalities implemented in the final Telematic Hub – amounts to 10 percent of overall claims expenditure in solution. Europe. In Italy, the issue is particularly bad. Unorganized and professional fraud is one of the reasons why Italy’s motor III. APPROACH insurance premiums have risen to the highest levels in the EU Setting up a brand new telematic platform is a complex (about 2-3 times higher than in France). This problem is even work. Even with appropriate investments it is very difficult to worst in southern regions of Italy where insurance frauds and quickly implement a brand new Telematic platform. Since tariffs are typically higher. Italy’s politicians have picked up on every system lifecycle consists of multiple aspects, including the topic and have demanded countermeasures from the business, budget and technical ones, the systems engineer has insurance industry. The Italian government set a requirement to create technical solutions that are consistent with the for motor insurers to install telematics devices in all vehicles business case and funding constrains. Having fully adopted the and use the data to help set premiums. SE Lifecycle Stages for this project, during the Concept Stage So, Insurance Companies in the past few years started to all the available documents in terms of RFIs, RFPs and market invest heavily in installing black boxes (OBU devices) in analysis had been collected in order to properly shape the high level customer requirements set while characterizing the IV. CONCLUSIONS solution space with a very early conceptual architecture. The Today the Telematic Hub solution is in Support Stage Concept Stage was crucial for refining the stakeholder needs as according to a continuous improvement approach. Through well as selecting the proper technology for supporting the obtaining data ownership, Insurance Companies adopted IBM solution. For example, by utilizing IBM’s Connected Vehicle Telematic Hub solution are able to understand how to improve Insight cloud managed service, the System Engineers were able the quality of Telematic data of trips and crashes with single to quickly identify the appropriate solution architecture for GPS point. They identify driver behaviors clustering driving building the brand new telematic Platform leveraging out of the styles. This creates the foundation for Insurance clients to meet box capabilities including map matching, trajectory data their goal of achieving more effective fraud detection. The management and analytics, driver behavior analytics, driving online detection of vehicle position (in the case of car crash) context analysis, mobility footprint profiling, trajectory pattern also allows Insurance Companies running Telematic Hub to analysis, map event pattern analysis, real-time and dynamic quickly react to first notification of loss, putting in place a map context management and map topology fencing. stronger anti-fraud system and a better risk selection. All the data engineered during the Concept Stage allowed Additionally, the IBM Team have been able to build a the Systems Engineers to make the decision about the proper component architecture designed around Connected Vehicle lifecycle approach to be adopted for the Development Stage. Insight SaaS, so Telematic Hub solution can today leverage During this stage the development team was focusing on Driver Insight provided analytics to build new analytics on top several activities about: of them. Those new analytics will ensure Insurance Companies running Telematic Hub competitive advantage while • Functional and non-Functional requirements; leveraging IBM technology covering all basic needs like: • Use Cases; • Support for specific devices designed for insurance • Context; case; • Interfaces; • Logistic services with body-shops covering all regions and provinces (Italy); • Component Model; • Mobile connectivity services; • Operational Model; • IoT Platform services; • Service Model; • Data Platform services and Infrastructure services; The Systems Engineering approach and practices have been key success factors for developing Telematic Hub for Insurance Industry as a real solution for production helping to meet customer specific requirements since the first provided input and for fully addressing the following nowadays main Insurance Industry needs: • Lower claim frequency and average claim paid • Become stronger at fighting fraud and better at-risk selection This led to a development project for implementing the required basic services: New telematic contracts, Voucher management, data lake service, analytics services, integrated management services, device management services, logistic management services, call center. The key gate for the Production Stage was the dynamic test REFERENCES performed directly with the customer to verify the correct [1] INCOSE, “Systems Engineering Handbook, A Guide For System Life behaviour of all the application compared to the business needs Cycle Processes And Activities”, Fourth edition, INCOSE-TP-2003- 002-04, and stakeholder requirements, activity typically named UAT 2015. (User Acceptance Test). The system went live for the [2] ISO/IEC/IEEE 29148 - Systems and software engineering — Life cycle Utilization Stage exactly after 9 months from the kick-off date processes — Requirements engineering. of Development Stage in perfect line with the project plan [3] ISO/IEC/IEEE 15288 - Systems and software engineering - System life cycle processes. milestones. [4] Falcone, A., Garro, A., and Tundis, A. (2014, September). Modeling and simulation for the performance evaluation of the on-board communication system of a metro train. In 13th International Conference on Modeling and Applied Simulation, MAS 2014 (pp. 20-29).