Making SENSE of Commercial Property Risk by Harnessing the Power of Digital Twins
Updated: May 7, 2021
As we noted in the first SENSE white paper, the commercial property insurance market of 2021 is ripe for modernisation. We quoted the disappointing 2016-2019 Lloyd’s property insurance results in our first paper and we now know that Lloyd’s property underwriters lost a further £2.1bn in 2020. This takes the total losses from property insurance in the Lloyd’s market to an eye watering £4.75bn over the last five years (2016-2020).
The status quo is clearly not working and this is bad news for the customer as insurers are forced to increase prices and restrict coverage in response to the mounting losses. The big problem with the underwriting model today is that insurers lack comprehensive real-time risk data to help them monitor, detect and mitigate commercial property risks.
We know that fewer than 5% of commercial property locations are physically risk surveyed by insurers due to cost pressures, and even when they do send an engineer on site, they are very poor at following up on risk improvement programmes with only 1 in 7 risk recommendations being actioned. This leaves both the client and insurer exposed to pick up the bill when disaster strikes. Therefore, it is no wonder that commercial property insurance is currently loss-making and we believe there is now an urgent need for rapid transformation to real-time
(Image © Intelligent AI Limited 2021)
Digital Twins to drive change and make SENSE of risk.
With the introduction of low-cost cloud computing, faster data processing and advances in artificial intelligence for data extraction and image analysis, it is now possible to accurately develop digital models of risk at commercial properties without having to visit them.
These ‘Digital Twins’ of risk, are made up of open data (e.g. crime, fire, rateable value, financial data etc.), satellite spectral analysis data, IoT (Internet of Things) sensor data, and real-time natural catastrophe data (e.g. river flow and flood data, temperature and weather data, rainfall, landslip and seismic data). They can create near real-time models to give a far more accurate 360-degree view of risk across 100% of commercial properties at a fraction of the current cost.
Digital Twins have the potential to enable insurers and brokers to
accurately analyse and price risk far more quickly and accurately than ever before
undertake desk surveys in a fraction of the time and cost
accurately analyse whole commercial property portfolios
promptly identify high and low-risk properties
track and drive-forward risk improvement programmes with commercial clients
aid portfolio steering to encourage a diverse mix and balance of risks within any given portfolio
Digital Twins help insurers identify and create more profitable business and bring down the total cost of risk by reducing both claims cost (frequency and severity) and operating expenses.
In addition, Digital Twins have the potential to enable commercial property owners to analyse property portfolios before they are acquired, to drive down risk and enhance both health & safety and business continuity, whilst allowing users to better understand the risks not only within their estate but also potentially right across their global supply chain.
The market is starting to see the benefits of Digital Twins, with Trevor Maynard, Head of Innovation at Lloyd’s of London stating the following at a recent Lloyd’s Lab innovation event:
“Now more than ever intelligent real time risk information is needed. Producing a digital twin of a company, that can then be used to explore risk mitigation methods [is] a truly great idea”
Additionally, AXA noted:
“As a larger insurance company, it is extremely difficult currently for us to get a 360 degree view like Digital Twins could deliver”
(Image © Intelligent AI Limited 2021)
Unlocking the data from Risk Engineering Surveys using AI and Digital Twins
As well as being able to deliver Digital Twins for commercial properties without the need to visit them physically, it is now possible to use AI to extract and intelligently analyse data from tables and text in risk survey reports and then use Digital Twin insights to deliver interactive dashboards to augment and enhance the survey reports and deliver enhanced real-time risk insight.
By way of example, a major food processing plant with a faulty sprinkler system that is not fit for purpose is likely to be unknown to their major supply chain customers despite the fact that a total loss of the factory and production would cause significant loss and disruption throughout the supply chain. If the supply chain is aware of this enhanced risk then they may be able to help in prioritising the risk improvement.
The current model may be broken but we believe the use of AI to better analyse risk reports, together with Digital Twins of risk across 100% of sites can help to significantly drive risk improvement by providing real-time dashboards to track and communicate risk and deliver risk benchmarking across commercial property portfolios.
Examples of how digital twin data can drive down risk and costs
Accurate address data
Many property portfolios contain insufficient or inaccurate location data. In one commercial property investment portfolio reviewed recently, containing 2,800 apartment blocks with over 70,000 apartments, only 40% had latitude and longitude data. Even then, some 30% of those were incorrect by up to half a mile, which is a considerable margin of error when looking at proximity to underground tanks, seismic activity and closeness to bodies of water or wildfire zones. Insurers incur substantial operational costs in manually correcting location data and unexpected losses from inaccurate risk models. Digital twins can deliver accurate lat/long data across 100% of the portfolio, reduce losses and lower operating costs by removing the need for manual data correction.
Parametric insurance solutions
Digital Twins can also deliver real value in the underwriting of parametric insurance products. Applying AI and Satellite image analysis across a commercial property portfolio can identify the building’s condition before binding the policy. Suppose there is a claim for roof damage on a factory caused by hail. Digital Twins can quickly analyse local IoT weather stations within a few miles of the site to determine if weather conditions on the day of the incident are consistent with hail. Furthermore, Satellite images directly before and after the date of the claim can be automatically analysed using AI image analysis and synthetic aperture radar from the satellite to determine if the claim is genuine and calculate the percentage of damage and automate a pay-out without the need to send a loss adjuster. This provides the client with a far superior customer experience, minimises business interruption, helps mitigate the insurer’s claim, and lowers operating costs.
Furthermore, Digital Twins and IoT data can be used to predict and prevent events. Increasingly, data can be gathered to create hyper-local short term weather incident forecasts and provide upwards of a 3-hour hyper-local warning of an impending event. A more accurate, digital twin generated flood model could automate alerts to the locations in that area that there is a very high probability of a flood in the next 3 hours. The customer could then move stock to higher ground and take other preventative actions to help prevent or, at least, mitigate a significant part of any loss and resulting business interruption.
Digital Twins can generate macro-level assessments far more quickly than current manual methods where a disaster has occurred across a wide area, such as the Hurricane Katrina induced New Orleans Floods of 2005 or the Beirut ammonium nitrate store explosion in 2020. These major disasters take weeks or months for insurers to understand the magnitude of impact/loss and focus their support. AI satellite image analysis and digital twins can provide live ‘catastrophe risk analysis’ and provide ‘cumulative risk estimations’ in days or hours and increase the support that insurers can provide to their customers and help speed up interim claims payments.
The Iceberg effect (insured versus uninsured losses) - Modelling the Total Loss
When considering the return on investment for risk improvements, it is important to consider both insured and uninsured losses. In the case of a major global corporation that experiences the loss of a key manufacturing and distribution centre, the physical damage may be insured including the resulting business interruption directly linked to the damage and the rebuild or reinstatement period.
However, any insurance policy only goes so far and what is never covered are the customers you’ve lost, the growth you’ve forfeited, and the penalty you’ve paid (and will continue to pay) as investors lose confidence in you. These losses can destroy enterprise value, if not enterprises themselves.
Insured losses are often dwarfed by the uninsured losses that can be a factor of 4x-5x
By way of a case study, we are aware of a client considering investing £1m in flood defences to defend property insured to the value of £190m. However, with uninsured losses, the company calculated that they could lose a further £450m from its market value as follows:
Customer Loss caused by protracted business disruption to cost the corporation £145m.
Loss of growth over the short and medium term, caused by the 24 month rebuild to cost the company £240m against its earlier projected growth statements.
Loss of investor confidence, caused by lack of contingency planning and poor risk mitigation, driving down investor confidence and resulting in higher cost of capital to cost the corporation a further £65m.
Most commercial customers are too focused on the insured losses and fail to factor in these uninsured losses in their ROI calculations when approving risk mitigation costs. Digital Twins can help both customers and insurers to drive increased risk mitigation.
Benefits of Digital Twins
Digital Twins, delivering real-time risk insight, have the potential to lower the cost of risk surveying from £750 to £5000 to less than £10-£20 per site. Actionable insight can be delivered in seconds or minutes rather than 1-7 days (depending on the size and complexity of the site). They can deliver a 360 degree view of risk on 100% of sites and significantly drive down claims and losses for insurers and commercial customers.
Better risk profiling