The Low Down
At the heart of the business is the loco fleet and the work they perform daily: pulling and pushing wagons between locations, carrying all types of loads, and delivering services quickly and efficiently. These assets are supported by drivers, ground staff and engineers who work round the clock to keep goods moving.
In addition to the locos, are the wagons and employees who operate and work with the assets, all organised by resource planners, who meticulously plan out what assets and employees are required to form trains to deliver the service to DB Cargo UK customers.
A shared vision for optimisation
At the beginning of our engagement, we worked closely with the Head of Digital Transformation to define our shared vision. We believed there must be a way we could use modern computational algorithmic techniques to solve the supply and demand challenge at the centre of resource planning and provide a technology feature that could optimise and drive efficiency in the resource planning process.
Our very own Dr Tom Helliwell is an advocate for cyber-physical systems and optimisation. His thesis “Reconfigurable Scheduling through Discrete-Event Systems” focused on solving the combinatorial problem found in manufacturing asset scheduling. He presented simulation-based techniques to solve this challenge, proving that general-purpose optimisation techniques are not efficient enough to tackle the combinatorial problem in scheduling, where a discrete set of rules must be applied. Tom and I spent time brainstorming how we could apply his proposed manufacturing technique to the rail scheduling problem.
In stage one of our optimisation process, we focused on the computational algorithmic simulation technique, which allocates ‘demand to supply’ staying within bounds of the restrictions rail paths put on the operator. The simulation highlighted the opportunities DB Cargo UK had to drive efficiencies through technology. Unfortunately, I can’t show you just how successful the algorithm was, due to the sensitivity of sharing such data.
The success of the simulation is apparent, but where do we go from here – what does the future hold?
At this point, we now must look to the future and what the vision for optimisation will look like. Returning to Tom’s thesis – considering cyber-physical systems and the manufacturing digital twin – the same concepts will be applied to rail. The rail industry is still behind other industries with respect to innovation and new technologies, so now is the time to change that. Now is the time to innovate. The rail industry needs a cyber-physical rail platform; the digital twin of the rail industry.
This platform would host the digital twin of the operator, integrating real-world data streams: events from network rail and loco edge devices including GPS, black box, sensor readings, weather information, asset maintenance information, and track maintenance information. A true digital twin representation of the rail operator surfaced through an informatics platform, which can react in real-time to business or operational change.
Where does optimisation fit into this cyber-physical system?
With the simulation, DB Cargo UK can efficiently and effectively optimise the performance of resources and assets with minimal white space. Simulating against the digital twin will allow them to understand the impact of outside influences, including new customer orders, asset and rail maintenance, and even weather conditions impacting daily routes. DB Cargo UK will be able to react in real-time to external factors outside of its control, whilst maximising the performance of its assets.
When all is said and done, I believe that optimisation embodies the bright and innovative future of supply chain logistics.