IoT data is the foundation of smart-home insurance through enabling insurers to optimally assess and manage certain risks in their policyholders’ homes, resulting in superior value propositions while providing value-added services.
Sensor Networks’s bespoke sensors document and contextualise a policyholder’s daily behaviour with regards to water usage, electricity consumption, home frequency, shower patterns, sleep patterns, holiday patterns, security alarm behaviour and other inside-the-home insights. The smart-home data ecosystem leverages the granularity and interaction of the sensor data with analytics and machine learning, to draw personalised insights about users.
Geysers, or hot-water cylinders (boilers) as they are known in other parts of the world, are commonly housed in the roofs of most homes in South Africa. There are an estimated seven million geysers in South Africa, with with approximately 50,000 being sold as replacements or for new homes. Geysers commonly have a lifespan of about seven years, and are the most common cause of a household insurance claim, due to the consequential water damage that results when they eventually fail.
In April 2019, a user with a smart geyser and smart security device was monitored. Figure 1 displays the aggregated data for the user over three days, beginning on Easter Friday (19 April 2019).
The Sensor Networks sensor-data ecosystem revealed the following insights over the weekend:
| Sensor Data: |
| The geyser tap reveals that at least one resident was awake by 07:00 and three medium length showers occurred between 08:45 and 10:00. This is derived from the interaction between the multiple geyser sensors and their consistency with the alarm sensors. |
| At around 08:00 the alarm partition state change from ‘Armed Stay’ to ‘Ready’ and immediately to “Not Ready” + subsequent water usage reveals that the house members are awake and active. |
| Water usage ceases around 10:00 and the alarm state switches to “Armed Away” revealing the house is securely vacant which correlates with the lack of water usage until the family’s return on the Sunday evening at 19:10 and preparing for sleep by 21:00. |
| Insights: |
| The sensor data reveals when the family sleeps, wakes up, take their showers, have supper and when they leave their home. |
| The sensors tracked the hot water usage, the amount of energy consumed, the patterns of consumption and the wastage during this period. |
| The number of persons in the households can be confirmed through water usage patterns. |
| User wasted 10 kWh of energy while away for the weekend. |
| The sensors reveal the time when the family is in and away from their home. |
| Home activity is monitored when away. |
| Heating Schedules set by the user predict the expected energy wastage while the user is away. |
Data Driven Home Insurance
The insights drawn by Sensor Networks are primarily a vehicle that helps insurers enhance their customer value-added services, loss control, risk selection, behavioural change and optimised risk-based pricing. An aggregated weekday behavioural view further pronounces this by defining the home’s character at different periods, as seen in in Figure 2 below.
The insurer is now able to leverage this data ecosystem to enhance homeowner profiling by understanding the home’s dynamic risk patterns for optimised risk selection, mitigation and providing services such as energy saving.
Over time, the Sensor Networks IoT data ecosystem builds an accurate historic, present and future looking personalised behavioural profile of the household.
Figure 3 demonstrates the vast range of statistical insights which can be drawn from a user’s personalised data to develop a personalised profile. This, enhanced with population statistics, enable Sensor Networks’s data ecosystem to draw further insights through correlation. For example, a power outage signal in the East Rand can be picked up in real time and action taken to prevent dry heating of all geysers in that area.
In 2018, this enabled Sensor Networks to have a real-time understanding of the usage pattern changes for Cape Town users during the drought, while also enhancing the ability to create user energy-saving profiles based on behavioural criteria. These are some insights which can be leveraged by insurers to enhance their value proposition in the Age of Data.