You may not have heard of MIS, but the general topic of using satellite imagery to price asset values, risk, cat events, or claims costs after a storm or hurricane event, is well established in the world of insurance. A clear view post-event is the bedrock of parametric cover for example. There are other benefits of collating ground level data, from Buildings, Commercial, traffic volume risks, forest fires and more of course.
To learn more about McKenzie Intelligence Services IE chatted with Forbes McKenzie recently.
Can you give us an overview of McKenzie Intelligence Services and the core services you provide to the insurance industry?
Harnessing the unique combination of trusted global data sources, the power of AI, and expert military/NATO-trained analysts within our Intelligence team, McKenzie Intelligence Services (MIS) provides insurers with immediate actionable intelligence around their affected portfolios, policies and properties in the aftermath of global catastrophic events.
The MIS approach ensures that insurers aren’t inundated with evaluating, interpreting, or purchasing data themselves. Following an event, with MIS as a trusted partner, relevant data from sources such as satellite/aerial imagery, radar, drones, mobile phone data, social media, and sensors is assessed by the Intelligence team, and in-depth reports are subsequently produced.
Using their disaster-response platform, the Global Events Observer (GEO), MIS uploads these reports and overlays them on the relevant locations with client data, covering potential exposure, likely incoming claims and building-level damage assessments to give insurers an accurate, detailed digital representation of the impact of an event, ensuring they can quickly and reliably respond whilst ultimately keeping their promises to affected policyholders.
How does McKenzie Intelligence Services utilise geospatial intelligence to support insurance companies in claims processing?
By having a team of military/NATO trained analysts with decades of intelligence experience, we can ascertain critical information from our various data sources about affected areas following a major catastrophic event.
Comprehensively assessing this data, the team will identify where and what may be damaged, the likely extent of the damage and the proximate cause of it. Once they have completed these assessments, they compile their findings and create a product to be published in GEO.
Taking this actionable intelligence and applying it to geospatial and client datasets in GEO then paints a picture of what is happening on the ground in the immediate aftermath of an event, where relevant locations, portfolios and properties that may have been impacted are highlighted so that clients can get ahead of potential claims.
Typically, clients will use GEO to inform their decision-making processes around claims, where the level of detail and accuracy provided by the Intelligence team enables them to make calls on providing full or interim payments to those affected without the need for time and resource-intensive adjusters.
Equally, with ongoing events like hurricanes, wildfires or war, the MIS team will regularly update GEO with the latest intelligence from our analysts, so clients can monitor any potential new incoming claims as an event evolves, as well as access further insights around how their portfolios have already been impacted once more data becomes available.
What specific technologies and data sources do you rely on to gather and analyse information for your clients?
We rely on a huge breadth of trusted global data sources to support us with our multi-peril, multi-region coverage of events as we look to provide clients with the actionable intelligence they need to respond effectively.
Typically, we use technology like satellite/aerial imagery, radar, drones, mobile phone data, social media, and sensors to provide us with the data we need to uncover vital details about an event’s likely impact.
Though this of course varies by event and the technology we need to respond to that type of peril, such as specific imaging formats to see through the smoke created by wildfires, like infrared imagery or synthetic aperture radar (SAR), we have a variety of sources to draw from to help us do this.
Imagery is a key component as it tends to contain the most important details about an event’s impact that our Intelligence team can observe, where their training enables them to spot minute details around how locations have been affected. This is usually sourced from our trusted imagery partners, who we have worked with for several years and as they supported us in our response to a wide range of global events.
Open-Source data like social media and news reports are also important for establishing context and clear up what is happening on the ground, where the team monitor various news sites and social media applications to find reports on events as they happen in real-time. This is then fed into all our products on GEO to keep clients informed of the latest developments.
Other important open-source data that we rely on includes tracking things like wind speeds or Maximum Estimated Size of Hail (MESH), where we use freely available data from dedicated organisations such as NOAA to help understand the likely cause and extent of damage in affected areas.
Can you share a case study or example where your services significantly benefited an insurance company in managing a major event or catastrophe?
Within 12 hours of the magnitude 7.8 earthquake striking Turkey and Syria in February 2023, GEO provided clients with a detailed heatmap of affected areas which pinpointed insured buildings, determining which of their customers had been affected and the likely extent of the damage to insured property. Using GEO enabled these insurers to quickly secure local resources, prepare their local claims team to respond to the event and make payments to affected policyholders on the same day.
How has the demand for your services evolved over the years, especially with the increasing frequency of natural disasters and other global events?
To understand how the demand for our work has evolved, it’s important to establish that before we began to provide value-added intelligence onto small satellite data sets, the market didn’t yet exist.
Much of the early work at MIS had involved travelling around the world on behalf of various governmental organisations, large corporations and insurance firms, where myself and our small team quickly developed a reputation for getting things done, no matter the location.
It was only after having garnered attention for our work in the London insurance market that we were awarded a grant to cover natural perils, which was highly successful and the start of our journey towards developing GEO.
The peril work led me to Lloyd’s of London, where alongside key heads of departments, I helped set up the Lloyd’s Strategic Claims working group. This was instrumental in developing GEO, where we benefitted from a direct feedback loop for GEO’s development from the market it was designed to serve.
Working closely with Lloyd’s, we effectively created a new marketplace for delivering this type of value-added intelligence to insurance providers. By agreeing upon a centralised contract with Lloyd’s, their managing agents became the initial market for GEO.
Then, in August 2021, GEO was officially launched as the central catastrophe intelligence platform for Lloyd’s, having recently added drone, SAR (Synthetic Aperture Radar), satellite and aerial imagery capabilities, which was an incredible achievement for everyone involved with MIS. Having the support of Lloyd’s from the outset was vital, as it enabled us to demonstrate our capabilities by serving the market we had helped to establish.
Once we had delivered a solution designed to meet insurers’ disaster response needs, the market was very much there and waiting, and with the increasing frequency and severity of catastrophes, the demand was immediately clear.
In the nearly three years since GEO’s launch, the evolving demand for our services is reflected in the scope of GEO’s growth, given that it now provides services to over 1,200 users across more than 70 clients, with over 80 million lines of client property risk and nearly 200 global catastrophic events covered, including hurricanes, earthquakes, wildfires, and war amongst many others. The number of GEO users and events responded to has consistently increased year-on-year.
As we continue to see insured loss records being shattered year after year and major incidents like Hurricane Beryl happening earlier than ever before in the hurricane season, I’ve no doubt that this demand will continue to grow, and that as we continue to develop GEO, we will be always there to solve insurers’ evolving challenges.
What role does satellite imagery play in your operations, and how do you handle the sheer volume of data that needs to be processed?
Satellite imagery is a vital aspect of our operations, especially following events where it is difficult to get boots on the ground or a plane in the sky.
Of course there can be limitations, such as cloudy skies impacting visibility, but satellites have a consistent cadence of providing imagery, so we can be aware of when it’s coming in and as a result, have our team ready to make assessments as soon as we receive this data.
Although the resolution of satellite imagery is typically lower than that of a drone or a plane flying directly overhead an affected area, for instance you might get a 30cm resolution satellite image as opposed to a 7.5cm resolution aerial image, the experience and expertise of our Intelligence team in analysing imagery means that they can still make highly accurate and rapid assessments of impacted locations despite the differences in resolution.
This is really the key reason why satellite imagery plays such an important role, as it enables us to do our job and provide the actionable intelligence to clients in the immediate of an event, without the need to rely on tasking a plane to capture aerial imagery, which can take days if the weather conditions are not suitable.
In terms of processing the vast amounts of data following an event, we have multiple ways of doing this. First and foremost, we rely on the Intelligence team to manually assess things like imagery and key data, where their training is vital to establishing ultimately what is relevant, and then go one step further to make important calls around what this data is telling us so it can subsequently be applied to a report in GEO.
Although the team is relatively small, their expertise allows them to perform these tasks with speed without sacrificing accuracy, which is a trait that we have applied to our AI-driven building-level damage assessments, another way we are able to manage the vast amounts of data at our disposal.
By training our machine learning models on real human-led assessments from previous events covered in GEO, with a human analyst in the loop to ensure the results are correct, we have been able to produce an AI model that can rapidly make assessments to a high degree of accuracy whilst covering a much larger area and amount of data than humans can in the same period.
Whilst every event creates a tremendous amount of work for the MIS team in terms of processing data, it is a challenge they rise to every time.
Over the next decade, what are some of the key trends or developments in the field of geospatial intelligence and its application to the insurance industry that you are excited about?
With the goal of enhancing disaster response as events become more frequent and severe, I am excited to see the new technologies that are constantly being developed to help uncover important information in the immediate aftermath of an event’s impact.
Being on that journey ourselves, we are always looking for the best data sources which give us the ability to quickly and accurately understand what is happening on the ground. So, if that’s improved imagery capabilities or new ways to capture geospatial data around events, we can harness this to evolve our offering.
From an MIS perspective, I’m excited to be working towards supporting more business functions within the insurance industry.
Equally, I am also looking forward to continuing the development of our recently released AI building damage assessment feature, which is trained on a vast amount of examples of human-led intelligence assessments. It is constantly improving by being trained on new data that follows each event we respond to, where our intelligence experts are always on-hand to ensure that it is meeting the expected standards.
Be the first to comment