Radius Payment Solutions Ltd hosted Brandon Kelly on industry placement

Published on

Radius Payment Solutions Limited is a fuel card company founded in Crewe in 1990. Since then the company has expanded and grown across Europe to become a global competitor who agreed to host LIV.DAT student Brandon Kelly for his 6 month industry placement, an integral part of the Centre for Doctoral Training (CDT) PhD scheme. The CDT PhD schemes specialise in their industry links relative to other PhD schemes and the industry placement is an excellent opportunity for students to apply their bespoke LIV.DAT training to the real world, while also gaining knowledge and experience of working outside of academia. While at Radius, Brandon utilised his training in Big Data management to automate aspects of Radius’ data processing.

The purpose of a fuel card is that a business which own feets of vehicles will require their drivers to refuel those vehicles on journeys. Rather than the drivers paying for fuel out of pocket and being reimbursed by the company, the fuel card allows the driver to buy fuel using a company fuel card. Another product Radius distributes is Kinesis, a black-box style tracking device which collects and transmits up-to-date data about the vehicles journey. Both these devices and the fuel cards are used across Europe and all of them feed data back to Radius, the textbook definition of Big Data!

The data fed back by Kinesis per journey includes vehicle ID, journey start time, journey end time, expected journey end time alongside a wide array of other features. Separate to the overall journey data, Kinesis also creates “events” throughout the journey. An “event” is created every time either certain things occur (such as engine off, engine on, harsh braking/accelerating etc) or a certain amount of time has passed since the last event. Event data includes certain features not stored in the data output from the overall journey data including fuel level, in millilitres and as a percentage, and “event id” which indicates what triggered the event. It is these features which Brandon was focused on.

By comparing the fuel level stored in an “engine off” event with that of the next “engine on” event, it is possible to determine whether the vehicle has been refuelled. By taking the vehicle ID from the event, the vehicle can be matched to a customer ID and therefore to a fuel card ID. By comparing refuel events to transactions on Radius fuel cards, Radius can determine whether that refuel was paid for by a Radius fuel card or through a competitor. For the benefit of the customers it is also possible to determine whether the amount of fuel bought matches the amount of fuel put into the vehicle. If this is not the case, the driver may be refuelling their own cars or jerry cans for personal use. While this comparison is simple enough for one or two vehicles, more substantial comparisons across the approximately 7 million vehicles across Europe with such devices require the kind of big data management techniques which CDT training focuses on.