Big data and the Internet of Things are two technological wonders that are changing our lives. Our machines, equipment and devices are continuously connected and communicating with each other, and billions of devices are recording data at all times – sensors, microphones and cameras are attached to everything that we have – whether that’s industrial hardware, our phones, our cars – or even in some cases our fridges!
Insurance is nothing new – but imagine how hard it must have been for insurers to know what was going on in the mid 19th century. Quantifying losses for business interruption insurance back then was hard because modern accounting and reporting standards weren’t yet a thing. Most records were created by hand – and even just a few decades ago, when computers were becoming more common, but data was transferred by the lovingly nicknamed ‘sneakernet’ – walking to a colleague’s desk to give them a floppy disk with the data – record keeping wasn’t up to modern standards.
Today, we store data about the data that we store! That’s not a hyperbole. Datasets are huge and we track as much as we can about everyone and everything that we deal with, and we need new ways of working to handle all that data.
BI insurance has not changed all that much over the last few decades –
But increased access to information is starting to shape the insurance industry.
Big Data and Loss Reduction
GE is using big data to monitor the temperatures, vibration and performance of wind turbines all over the world. They get real-time, constant information which can be fed into Predix – some proprietary software – to help improve the efficiency of their operations and to make it easier to identify hardware that is at risk of failing. It allows for enables preventative maintenance before units malfunction, which can reduce the number of claims they need to make, and reduce downtime.
Big Data for Underwriters
Car insurance companies already use telematics to identify risk factors. Why don’t other insurance companies do the same thing? Well, to an extent, some of them are starting to find ways to identify risks and use that to predict who is likely to need to claim when people compare insurance.
Settling Claims Faster
When a large energy claim is made, the files can be hundreds of megabytes, or even gigabytes, in size. The models that are needed can require complex programmes and will look at everything from refinery linear programming to petrochemical production data, power plant data, and more. Performing ‘forensic’ analysis on this, and doing Enterprise Resource Planning to calculate BI losses and other data is a complex job – but could well save time compared to trying to calculate claims by hand.
It is more efficient to do this analysis and to work on issues of making interim payments and or calculating the final settlement figures – while then using the data from this claim to inform future policies and claims.
Measuring Loss Accurately
The more data we have, and the more granular that data is, the better equipped we will be to calculate factors such as heat rates, fuel burn, turbine efficiency losses, and the impact of the non-catastrophic damage. If we can identify the loss of capacity on a specific turbine, or understand the availability of hardware during a partial outage, this can help us to understand the interdependencies of different units, and can make loss calculations more accurate, and easier to do.
Big Data and Cyber Security
One issue with big data is that of cybersecurity – if a computer system is breached, then it could leave that business under threat – either for trade secrets or for loss of control of their systems if remote management is an option.
In Conclusion
Big data has some way to go before it becomes user-friendly, but it is a useful tool, and if we improve the collection and use of the data it could mean a lot organisations.
