11 Mar 2021
The Texas Snowstorms have highlighted the desperate need for AI in disaster recovery
By: Jimmy Spears, MAL, MSPM, AIC, AAM, AIS
North America gets its fair share of snowstorms. However, Winter Storm Uri was unlike anything we’ve ever seen, leaving over 73% of the Lower 48 covered in snow.
Of all the states affected, Texas was hit particularly hard, with temperatures dropping down to 3° F. That’s lower than the average temperature of Alaska at this time of year.
Texas has its own power grid, meaning that when the coal, wind turbines, natural gas pipelines and water mains all froze over, the state was left to fend for itself. Unlike disasters like hurricanes that tend to be isolated to specific areas, this is one of the first times that all 254 counties in the state of Texas were affected by one incident. This meant that millions were suddenly without power or water, some for up to five days.
Demand for electricity jumped as people tried to heat their homes, and this led to the introduction of rolling blackouts to ration energy. Many even left their faucets dripping in an attempt to stop their pipes from freezing.
What’s the damage?
First and foremost, we must acknowledge the fact that these storms did result in the loss of life. At the time of publishing, dozens of people were sadly killed and that must not be forgotten or diminished.
Economically speaking, we could be seeing up to $295bn worth of damage in Texas. To put this into context, Hurricane Katrina came in at a record $161bn total.
The biggest bill will be for residential and commercial property, mainly from water damage and flooding caused by burst pipes. On top of that, there will be snow and ice damage to the outside structures.
Businesses will have to manage the effects of the power outages and inability to trade, on top of any structural and stock damage experienced.
Shocking! 😱 Turn up the volume!
Pipes busted from freezing. (Denton, Texas)
Please, If you are in Texas and you have not done so, let a couple of faucets drip until the freeze is over! #TexasFreeze #Texas #Denton pic.twitter.com/PE9Fy9jv7w
— Anas Alhajji (@anasalhajji) February 15, 2021
The storm has also led to at least $600m in agricultural loss, from the destruction of crops and livestock as well as the overall impact on operations. This will have long term implications, as it’s not as simple as losing one season of crops or produce. It will take years before the damage is reversed.
And of course, the snow and ice caused problems on the roads, leading to accidents.
The road to recovery
Normal property damage claims, like a burst pipe, are usually taken care of on the same or next day. This is standard for the industry and what policyholders have come to expect.
However, these storms caused a massive surge on resources, from water mitigation teams to adjusters to the vendors themselves. There simply aren’t enough bodies to go around and help everyone. The New York Times put it well: ‘Plumbers were suddenly like roofers after a hurricane. Everybody seemed to need one, all at once.’
Effectively, homeowners have had to fend for themselves and mitigate the losses they’re facing while they wait for help. In general, this involves trying to make houses habitable again – hopefully avoiding expensive hotel bills (or even being forced onto the streets) while also preventing property damage from getting more severe, and expensive to repair, before it can be assessed.
A study by AIR Worldwide suggests that the average damage for residential structures comes in at around $15,000, and for commercial buildings it’s $30,000. Any way to reduce this is beneficial for both the consumer and the insurer.
How can we prevent such devastation in the future?
We can’t control the weather, but there are two important actions we can take: (1) prevent these events from causing such severe damage in the future, and (2) speed up the response time to ensure help is given when and where it’s needed.
Identifying areas of high risk
On 1), we can better identify areas that hold the highest level of risk. Once you can identify risk, you can better predict loss and find ways to reduce the potential impact, for example by weather-proofing buildings.
One way to do this is to take thermal images of homes throughout the year and measure the thermal loss they experience. I would wager that the homes with the most thermal loss were the ones that suffered the most, as their pipes likely froze and burst before better insulated homes, leading to water damage.
Technology can also be used to look at various piping layouts and alert people to the areas and junctions that are prone to freezing. Work can then begin on making these structures stronger, and more able withstand lower temperatures.
Identifying those in greatest need
On 2), i.e. speeding up appraisals, there’s a real need to automate appraisals. AI can be used to quickly and accurately allocate resources at scale, helping those who need help the most.
I’ve previously spoken about Straight Through Processing (STP), a way to streamline data sharing in a fully automated manner, without manual intervention. This is something we’ve already achieved in the automobile industry. A person who has an accident can take images of the damage and AI is then used to assess it. The carrier can then approve and pay that claim right away.
There should be no difference in the ability to identify damage to a home. The right machine learning algorithms can be used to carry out an appraisal and provide the insurance carrier with a list of damages, meaning we can then identify where the quickest impact can be made, assign resources and mitigate the most loss.
Technology is by far one of the most effective ways to help us recover faster from natural disasters. No, we can’t prevent them altogether, but we can certainly lessen the impact.
Head over to our podcast series ‘Accident and Disaster Recovery’ to hear more about the impact of the Texas snowstorms and how AI can help people recover faster.
Want to know more about the AI technology I’ve spoken about and how it can accelerate P&C claims? Contact us for a demo.