Friday, June 14, 2024

Vehicle telematics and AI Simulators

I remember vividly back in the 80’s when working as a Transport Operations Manager, I received a call telling me to be ready for a helicopter landing in our yard.

One of our cargo trucks had been hijacked and our organisation had recently, on the advice of our insurers, installed tracking and tracing technology in our trucks. My job was to act as a spotter from the helicopter as the pilot navigated his way toward the beacon signal from the truck, in an attempt to secure the truck and cargo in coordination with the police on the ground.

This was my first experience with early vehicle tracking. The evolution from track and trace to advanced telematics has been a journey of discovery over the years working with various providers and their respective software and hardware solutions.

Today vehicle telematics measure everything from vehicle engine diagnostics to cargo conditions and driver behaviour on the road. This has become a standard amongst many companies in a move towards cargo visibility, route management and optimisation of assets, while maintaining a “Duty of Care” approach to customers.

A key area of interest for me, was how we could measure the drivers on road safety and driving performance without doing in cab driver evaluation, being as this takes a lot of time and resources given the fleet size.

Whilst speed, harsh breaking and acceleration were measured via accelerometers in the onboard devices, we still needed a common and consistent set of performance metrics to identify the outstanding drivers compared with those that needed training and development. Working with our solutions provider we established driver scorecard reports that identified aggregated behaviour metrics and allowed us to focus training on groups of drivers in an effort to close the performance gap.

At this point we had a feel for “the good, the bad and the in between”.

I was excited to learn about a new development in driver evaluation from our solution provider, who had teamed up with an Artificial Intelligence Driver Simulator start-up. After further research, we realised that the company had spent years working with various universities in the US, UK and EU along with OEM’s and racing technology companies to create an AI Driver Simulator, that through the use of biometric devices and a digital twin of Dubai, our drivers home city, we could measure highly detailed and individualised driver performance.

The results were a revelation as we moved away from aggregated data, often disguising on road incidents, towards driver specific behaviour. With the use of biometrics and the AI simulator experience, we were receiving an in depth and personal view about drivers who tended toward aggressive behaviour, distractions, speed habits, nervousness and many other attributes that are difficult to measure outside of an ongoing daily in cab evaluation.

It was incredible to think that a simulator could provide information regarding skills, personality, mental fitness, critical thinking and road rules compliance.

The data received has allowed us to individualise and hyper focus our training on specific behaviour and concerns identified by the AI Simulator, in addition to the general concept training programs.

Drivers continue to be monitored via the vehicle telematic scorecards for ongoing performance measurement, with the ability to send them to the AI Simulator for deeper evaluation should they drop below the defined performance KPI’s or receive a penalty from the traffic authorities.

We have seen an increase in performance, a reduction in penalties and most importantly a decline in accidents. As a result, we have adopted the use of the AI Simulator for all driver recruitment as a second layer of evaluation post interview and road test, allowing us to recruit drivers who have the right characteristics for the job before putting them behind the wheel of a truck and customers cargo.

There is no doubt that technology will continue to evolve exponentially to a point where driverless vehicles become the norm, however the journey will take many years before government authorities and the general public accept driverless mobility solutions as common place. Until then we have an obligation as road users and logistics providers to operate in the safest and most efficient way possible in our journey towards VISION ZERO.