Driving with Data: How AI is Personalizing the Auto Insurance Industry and Saving Lives
In this special guest feature, Gilad Avrashi, CTO & Co-Founder at MDgo, believes that now is the time for the insurance industry to leverage AI technology to unlock the power of data and provide personalized services that customers demand to retain them as loyal customers. As CTO of MDGO, Gilad spearheads MDGO’s technology vision and direction. Prior to founding MDGO, he served for six years at Rafael Defense Systems at the National Acoustic Lab, where he headed a team of engineers and researchers and conducted full-scale trials at sea. During his time at Rafael, he also headed the Algorithmics team of several top projects, received 3 Certificates of Merit and was awarded the Ministry of Defense’s prestigious Katzir Grant. Gilad holds a MSc in Electrical Engineering, Signal Processing and Communication from the Technion – Israeli Institute of Technology.
Data is transforming the way we live and how consumers engage with service providers. Most industries are using data to personalize offerings for customers and provide real-time expertise to strengthen relationships while gaining consumer trust. We are seeing this with Netflix making personalized movie recommendations based on the viewer’s specific interests and history. Now is the time for the insurance industry to leverage AI technology to unlock the power of data and provide personalized services that customers demand to retain them as loyal customers.
In addition to leveraging data for personalized services, insurance providers must compete on price to retain today’s customers. AI technology gives insurers the data needed to streamline the claims process, while reducing costs to pass savings off to customers. Leveraging data from advanced technologies to better understand drivers and accidents drives customer retention, brand loyalty and savings.
Despite the obvious benefits of leveraging data, very few insurers currently do it, creating gaps between their capabilities and customers’ immediate needs and expectations. For instance, according to the OECD, 44% of car crash fatalities could have been prevented if real-time data on the type and severity of their injuries were available for emergency medical services and treating hospitals. Indeed, AI technology has the power to uncover real-time insights from car accidents including occupants’ injuries and vehicle damage, providing lifesaving information to medical teams.
Utilizing advanced machine learning algorithms, ADR technology (Accident Detection and Response) can analyze the crash pulse recorded by vehicle telematics devices and virtual sensors to provide detailed reports. Beyond life-saving capabilities, AI can provide personalized services to drivers, such as automatically recommending a tow truck or autobody shop after an accident, as well as other location-based associated services.
ADR, powered by AI technology and machine learning capabilities, has huge potential for data applications in the insurance industry, and here’s a bit about how it works, beginning from the moment a crash occurs.
Gain access crash and injury data powered by machine learning
ADR’s initial, and potentially most significant, application of AI is to extrapolate crash test data to better understand and predict the connection between vehicle acceleration (or deceleration) and potential injuries. Advanced machine learning algorithms are used in tandem with physical sensor devices installed in the vehicle to accurately detect crashes as they happen. In order to address one of the most significant pain points of insurers, i.e. accurate and real time first notice of loss, which is driven by a low false negative and false positive rate. The physical sensor “wakes up” when it senses a shift in the vehicle’s acceleration, guaranteeing that no event goes undetected and lowering the false negative rate. Then, machine learning algorithms gain a complete picture of the crash, which makes it possible to remove the possibility of false positives. Finally, once a crash is detected, the damage and injury insights are provided using deep learning models, making it possible to deliver insurers detailed crash reports in real-time.
How is this done, you might ask? That’s where the concept of ADR’s unique virtual sensors come into the picture. Using cutting-edge deep learning networks, vehicle motion sensor readings are translated into virtual sensors, or a grid of dozens of sensors that are usually set up to determine potential vehicle damages and occupant injuries in crash test scenarios, such as those conducted by automotive OEMs. These sensors are virtually placed in and around the vehicle and its occupants, thereby emulating a world where you potentially walk around with an accelerometer inside of your head to assess potential injuries, without ever having to install any additional hardware. This imaginary “hardware” or virtual sensors makes it possible to determine where an occupant involved in a crash has been injured – delivering this information both to their insurer and, more importantly, to first responders and medical personnel.
Finally, each crash is identified by a unique crash pulse, which is used as its fingerprint. Machine learning algorithms analyze the fingerprint and construct a full crash profile, which is then translated into accurate damage and injury insights, providing an objective view of the accident. Access to these detailed data could decrease costs for insurance providers as gain access to a complete picture of a crash and its immediate consequences for the occupants.
Making the delivery of personalized services and care for policyholders possible
Beyond lifesaving capabilities, AI can automatically assist drivers in the recovery process following an accident. Having crash data readily available allows insurers to automatically make recommendations for recovery. Imagine receiving a message from your provider immediately after a wreck letting you know a tow truck is on the way and directing you to the nearest autobody shop. These types of personalized automated engagements are possible with robust data collection and processing beginning at the moment of impact.
An objective view of accident data is a critical and hitherto limited area of visibility for the auto insurance industry. ADR removes uncertainties by detailing the accident journey in data-driven reports for the insurer. Personalizing recommendations based on data could save insurers money during the claims process as they guide customers to supported garages and gain a clear picture of the accident the moment it occurs, which also delivers personalization value.
In all, AI has immense potential to transform the insurance industry as we know it. We’re about to witness a true top-down shift, be it from offering tailored, personalized services that begin at the moment of impact, to reducing claim costs for insurers by directing policyholders to engage with an affiliated network of services (i.e. DRP body shops and tow trucks). These capabilities and more are made possible with ADR’s real-time sensor data and custom-built AI algorithms, delivering fully objective reports with zero human intervention that help insurers deliver unique life and cost-saving value for policyholders.
Sign up for the free insideBIGDATA newsletter.
Join us on Twitter: @InsideBigData1 – https://twitter.com/InsideBigData1
Read More …