Edging towards a better UBI experience

In order to keep pace with the vastly increased data processing and storage demands of the connected and autonomous vehicle, many organisations around the world are now considering moving beyond a sole reliance on cloud technology and devoting more and more attention to the current and potential applications of edge computing.

Exploiting data

In recent years, Edge computing has emerged as a key tool in the ongoing push to expand the practical frontier of computing applications, data collection and services further away from processing nodes based in central locations and towards the extremes of a given network. In doing so, the approach is enabling a growing number of companies to analyse and exploit data collected at its source.

Although edge computing has, to date, tended to be used for fairly pedestrian uses such as processing, storing, filtering and transmitting data to cloud servers, many observers now believe we have arrived at the moment where such systems are now sophisticated enough to viably gather and interpret data collected at the ultimate location. Among other things, this is opening up the possibility that forward looking insurance companies might use data collected in connected vehicles in devising newer and ever more innovative products and services and derive ever greater value for themselves and their customers.

According to Paul Glynn, CEO at Californian software company Davra Networks, Edge computing is “by its very nature” designed to bring the most value in remote and mobile environments, so it should be viewed as no surprise that it should offer what he describes as “specific benefits in the area of vehicular UBI”.

“Processing data in the vehicle allows insurers to make faster, more dynamic decisions in relation to risk while also providing a stronger chain of evidence for claims,” he says.

Understanding vehicle usage

When it comes to providing services to the owners of enterprise and industrial vehicle fleet owners, Glynn explains that Edge computing can also be used to gain a clearer understanding of vehicle usage while in high risk locations a particularly useful attribute for mining, forestry and utility companies. He also argues that data gathered using the technology could come in useful when vehicle owners are carrying out higher risk tasks, such as when operating truck mounted cranes, pumps or dredgers, or even when vehicles are being operated by less experienced drivers. “In situations like these local data can be used to drive dynamic pricing models and offer more flexible cover to customers,” he says.

When it comes to monitoring remote assets, the collection and storage of large amounts of data can often be an extremely expensive, not to say inefficient, strategy. In helping to overcome such limitations, Glynn argues that Edge computing technologies allow both insurers and customers to keep their data within the vehicle but tag it for retrieval at a later date if required.

“This is extremely helpful in relation to video data, which is particularly bulky and expensive to transmit and store and can generally be managed ‘at the edge’ with a basic rules engine and video analytics platform,” he says.

Machine learning

Moving forward, Glynn highlights the regulatory environment relating to privacy issues as the key driver of any emerging hybrid Edge/cloud data model to enable distributed data processing with personal or user definable data staying within the vehicle and anonymized or averaged data being shipped back to the cloud. Speaking from the perspective of a company that is focused primarily on enterprise, as opposed to consumer, vehicles, Glynn also views artificial intelligence (AI) and machine learning as enabling technologies for new business models, particularly in relation to dynamic or point pricing offerings. “Having a vehicle model its own usage patterns and potentially change insurers on demand at different stages of a journey or task is extremely interesting to our customer base and could be an industry changing initiative if designed correctly,” he says.

“Our customers need to understand that they are reducing or offsetting their risk in the most efficient way possible at all times, so having each vehicle defining its own insurance requirements and dynamically changing them over time would be invaluable to them,” he adds.

[Ins.Williams.2017.11.22]

Connected Vehicles