NEWS

If as a fleet, you have multiple fleet risk suppliers producing multiple reports in multiple formats at multiple times of the year, you will probably appreciate the challenges of each stage of the image shown above.

Driversure absorbs raw data direct from many types of fleet risk service supplier and its systems are designed to remove the burden of managing BIG data. Without such technology, this element of the process represents heavy-lifting for fleets. Our programs make the whole process quicker, more accurate, transparent and, well just easier.

ANALYSING FLEET RISK DATA

We start by allocating all gathered data to individual Driver Safety Records. Once allocated, the data is tested against a matrix of maximum risk thresholds. Any data that exceeds the standard alerted and a review is carried out to validate and assess severity. At Driversure this transformation into SMART data occurs within the blink of an eye.

The number of data sources used has a bearing on the quality and accuracy of conclusions drawn, which is why we seek to include the four principal sources as a minimum (licences, claims, online assessments and telematics). Of course, additional sources of risk data would only add to the intelligence gained and the transformation ultimately achieved.

TRANSFORMING FLEET RISK DATA

A magnificent transformation occurs when analysed data is used to improve decision-making, which in turn leads to reduced overall risk.

Of course, this implies that you need a large enough pool of data from several sources, that the data is accurate and full and that the algorithm used to assess historic risk severity leads leads towards the most likely markers of future risk.

In other words, Driversure seeks the correct conclusions to deliver the correct remedial action to the correct drivers within correct timescales to achieve the correct outcomes

MACHINE LEARNING

Driversure now models the risk profile of drivers who have had accidents using more than ten years of gathered risk data for more than 10,000 drivers. This process allows us to teach our machines to spot similar profiles of those drivers who are yet to suffer an accident, so we can intervene to prevent such an occurrence in the future.

The process of creating this form of artificial intelligence has begun and will only improve further as our database grows and our algorithms are fine-tuned