The 700+ years of Insurance industry have seen three major eras: A manual era (15th Century to 1960), the systems era (1960s to 2000), and the digital era (2001-20X0)
The most common thing in all three eras is that the fundamental insurer’s business has been firmly relying on data analytics all through (though the volume and the purity of data have been significantly improving over the decades). However, the main difference between the eras is the speed of changes in adoption to leverage technologies to better the way risks are assessed and keep their capital intact.
While spectacular increase in data volumes (thanks to telematics taking the lead), insurance actuaries and experts are spellbound; the main challenge is the existing analytical models and algorithms being inadequate in advanced analysis to support insurers, which possibly could be done best only by machines.
The digital era has seen technology supporting customer engagement, delivering processing capabilities, providing single view of customer, add-on selling, analytics, etc. The next wave of the evolution will obviously be towards Machine learning & AI (also on the top of Gartner’s Hype cycle). This is pedicted to bring in much larger transformation in the insurance services business landscape and the way insurers do business.
The Bigger Picture: Machine Learning shifting Insurance Paradigms
Machine learning is not new. In fact, research for the infamous model, ‘Neuron Network’ was conducted during 1965-70, which did not kick-off due to the reason that processing power was not enough to effectively handle the long runtime.
Most of the insurers have been following supervised learnings for decades – illustratively means assessment of risk with known parameters applied on available information (structured data & unstructured data) in different combinations to get the desired results.
New-age insurers are striving to get to unsupervised learnings, where pre-set goals are defined; if there are changes in the variables, the system recognizes the changes and tries to reset according to the set goals (e.g. the GPS map suggests alternative routes dynamically based on traffic conditions). In the insurance industry, this learning is adopted for usage based insurance - Telematics/IoTs are early examples.
Reinforcement learning primarily depends on Artificial Neuron Network, in which the target/goal will change dynamically depending on objective. According to the environment, the variable will self-adjust to the dynamic targets/goals (e.g. driverless cars). The insurance world has not yet adopted this style.
Ultimately, the advent of technological innovations such as clothes carrying out pattern recognition and transmitting signals, underwater transportations, neuromorphic chips, 3D/4D printing and holograms, insurers are only going to have more opportunities to explore and transform their business models and ways in which they do business in future.
Machine Learning and Big Data Analytics to Revolutionize the Insurance Industry.
Determining the price of insurance is based on possible occurrences of claims. Machine learning can consider the behavioral pattern of the insured person/objects across different sources in similar segments and simulate the illustrative behavioral pattern to assess and prescribe the price of future claims.
Machine learning can help insurers to get the future behavioral pattern of the particular segment of the clients using the products and provide ways to collaborate with OEMs in determining the price.
Additionally, machine learning offers real-time assessment of price using feedback loop of claims record and dynamically changing behavioral patterns. For example, machine learning can provide spot discount on premium when the road traffic is very low.
Machine learning can help the insurers understand the exact intensity of the accident to assess the loss, which is currently not possible. This will be based on the inputs from IoTs, monitors, level of damage, etc. Machine learning can help insurers save huge cost in loss estimation.
Machine learning primarily needs identification of object, human, and recorded activity to apply algorithms so that it can monitor and detect fraud in real-time. For example, one of the most critical fraud practice challenges that insurers face is dismantling the vehicle, selling the parts, filing a ‘car lost’ complaint, and getting insurance coverage. If objects, humans, and blockchain technology are integrated into machine learning, the fraud can be detected even while the car is getting dismantled and help in preventing loss and associated cost to detect.Culture Change
The resistance to machine learning within the industry is multi-faceted – starting from the underwriters and the IT teams to operational management. The various reasons include:
- Loss of jobs
- Focus on current issues like building the scalable foundation for their current journey
- Non-availability of resources and dearth of required skill-set
- Financial constraints in form of funding the development/POC/research stages
- Insurers want clear evidence of success before adoption
- Regulatory restrictions and privacy norms
- Large diversity of training for real-world operation and simulation of various scenarios
- Initiation challenges
- Infrastructural challenges faced by a neural network designer include filling millions of database rows for its connections
While machine learning has the ability to prevent fraud during the moment of happening, current machine learning algorithms can certainly determine the likely fraud cases based on data and behavioral patterns. For comprehensive leverage of machine learning, HCL’s end-to-end service portfolio covering IoT, machine learning architectural framework, insurance experience, block-chain expertise, global application services and infrastructure leadership could be an ideal to a new beginning new-age connected insurers.