David Gallimore, Fleet Operations’ Chief Information Officer, looks at the role that AI could play in the future of the fleet sector.
Much is being made of the potential for Artificial Intelligence (AI) to disrupt the fleet industry.
But a word to the wise – in some cases, the use of this term can be misleading, and a healthy degree of scepticism is called for among fleet buyers.
Fundamentally AI can be defined, in simple terms, as the ability of a computer to make a decision and so it can be something of a broad church. It encompasses both data processing and task automation – either algorithm-based data reconciliation or processing and reporting underpinned by machine learning engines.
Caveat emptor: ask the right questions
Consequently, buyers should make sure they ask the right questions of vendors whose solutions purport to offer AI as a key feature or USP.
What type of AI have they got? Where in the platform is it? What role is it performing, and does it appropriately meet your individual requirements?
As AI moves away from advanced rulesets and algorithms and into machine learning, the quality of the AI’s implementation and available dataset becomes paramount. The better the quality of implementation and data, the better the decisions the AI will make.
For example, a decision on the ‘best’ vehicle to buy could be made solely on leasing cost data, however if service, maintenance, repair and fuel costs are factored in then a truer TCO comparison can be made. This decision can be enhanced even further if the AI can factor in a driver’s annual mileage and driving style, and then compare this with the real life TCO of other vehicles on the road.
Where multiple fleet data sources exist – AI may have an important role to play in significantly improving decision-making. Outsourced fleet management companies that sit across various fleet services, and that consequently have access to multiple such streams of data, are especially well placed to leverage tech developments in this space.
Currently AI is being used to enhance existing operations, such as predicting end of term for vehicle re-contracting, for example, by using AI to cross-reference with available re-contracting deals and mileage predictions. Or from a risk-management perspective, multiple data sources, such as behavioural data from telematics, a driver’s license data and accident history, can be analysed using AI to predict risk and to better target interventions through automated training and alerting.
A glance to the horizon
Looking to the near future there is also huge potential to significantly reduce TCO by helping to pre-empt service, maintenance and repair (SMR) needs within fleets. For example, if AI is combined with technologies such as ‘connected vehicles’, vehicle inspection data and an integrated supply chain – it’s easy to imagine a scenario where the fleet effectively self-books for SMR (for example when an engine warning light is displayed) pausing only to confirm a choice of appointment with the driver. This would effectively transform unscheduled maintenance into scheduled maintenance. AI could then be used to streamline technical and invoice validation, and approval, by automatically highlighting transactions which sit outside of the expected.
AI is also likely play a large part in the up and coming ‘mobility’ space where it may be used to compare alternative forms of transport and factor in employee entitlements and budget constrains to find the most appropriate form of travel for each leg of a journey – all in real-time during the booking journey.
While a healthy degree of scepticism is advised in the current market, buyers should be clear, however, that technology in this area has improved significantly in recent times and if AI is implemented correctly, they are investing in something truly innovative – not just the latest iteration of a technology that’s been available for many years.