26 June 2017 07:00
Greg Gilbert | Photo : Réjean Meloche
Even if an insurer is better prepared to cope with a natural disaster, if it doesn’t keep its promises in case of a claim, its efforts will be in vain, » said Greg Gilbert, chief scientist of the data in OPTA intelligence informationnelle.
Mr. Gilbert made the remarks during the Day of the insurance of damages 2017. He gave the example of the large forest fire that occurred in Fort McMurray in 2016. The claims adjusters have not had access to certain areas for periods ranging from four to six weeks. OPTA has used the picture of satellite to help with the work of claim. Two days after the passage of the flames on the city, images have been analyzed to better assess the extent of the damage.
Artificial intelligence systems have been used to target areas where we could confirm the devastation. According to the GPS positioning of these images, the data were transmitted to the insurer so that it can match this estimate to each client and to insured property in the disaster zone.
«Insurers sometimes have to react very quickly to situations. The people who were evacuated and who were forced to abandon their property waiting for responses from their insurer, » he says.
Immediately after a large disaster, insurers show interest in predictive models, but it does not last always, observed by Mr. Gilbert. The difficulty remains the fact to integrate the data into their pricing system. There is a lot of competition in the market. It is not always easy to impose additional invoices to the insured, he says.
«At every major disaster, we see it, there is always an insurer that is more affected than the other, because the risks were more concentrated in the region affected, and this risk exposure has been poorly managed,» said Mr Gilbert.
Insurers have tools to mitigate their risk exposure, » he said. It can mitigate its risks in three ways, » he adds.
First, the insurer must understand the risks that it covers. For each residence, he must have precise information on the year of construction, size and value of the property. He must also know the risk of flooding or sewer backup, plumbing issues or a pump, and the distance from the fire station closest to you.
The second point is the use of geomatics to conduct simulations. Various simulation models exist in the market, in particular on the risk associated with floods. Other systems will be launched in 2017, affecting the risk related to fires in forest areas.
The third point affects the management of the portfolio of the insurer. «If you are the insurer is dominant in an area vulnerable to flooding, you want to know how many properties are located within the flood-prone area «, he says.