Benefits for the business as a result of delivering this solution include:
• Provided the ongoing ability to quantify the primary transport cost impacts of alternative network designs with feasible routing strategies
• Greater visibility and understanding of operations and capacity across the complex 50+ depot network
• Established a new primary network planning team with integrated optimisation and simulation tools with processes to support future business growth and development
Background to the project
UKMail’s requirement was to re-shape their existing network though process improvement and the implementation of intelligent software tools. High primary network costs and preparing for a future of growth were the key drivers for this project. The software tools would be used to replace existing manual line haul improvement methods, inadequate for UKMail’s complex multi-site network. We partnered with Flo Group, specialists in logistics, transportation and change management, where our role was to provide the Network Performance Model as part of the intelligent software tools. The software tools enabled optimisation of the network routes, analysis of the operation through simulation and refinement of the operation taking into consideration all relevant legislation.
The software tools that make up the solution can be split into three main categories:
• Tactical simulation using the Network Performance Model (NPM)
• Transport optimisation using Llamasoft Transport Optimiser
Based on a given network and package volumes, the Llamasoft network tools optimises the routes based on given variables to find quicker and cheaper alternatives. The resulting routing rules are then loaded into NPM.
NPM uses simulation to model the network. Rather than using ‘off the shelf’ simulation tools, we developed a bespoke simulation engine. The main reasons for this are that it:
• Leverages industry standard Microsoft development tools providing access to a wide skill base and promoting supportability.
• Focuses on the specific requirements of UKMail and models the network operation explicitly
NPM is based on a relational database containing all the detail that makes up the network and the operation of the network from sites to vehicle schedules. The simulation model uses this data to run a 24 hour period of the operation, simulating each parcel that arrives at each site and traveling to its destination. A number of metrics are obtained during the simulation and reports are accessible on completion.
The transport optimisation phase takes an output from the simulation model detailing each vehicle movement including origin, destination, freight volume and distance. This data is then used to optimise how the freight is transported e.g. combine loads from two vehicles to make a single vehicle movement.
Solution in Detail
The NPM consisted of three main functions; data management, simulation execution and reporting. Through the data management function, all aspects of data the simulation requires can be imported from Excel spreadsheets, edited and deleted. The importing of data includes rigorous validation to ensure no data errors are present when running the simulation. One of the key aspects of the data structure is Scenarios, this gives the user the ability to redefine the data multiple times to test changes to the operation. Below is a diagram to illustrate NPM;
The simulation element of the NPM uses production standard Microsoft development tools. A Scenario is selected prior to starting the simulation, which defines the data to be used and what to link the reports to. The simulation has three main stages, where the last two are repeated until all freight is delivered or the end of the day is reached.• Sorting of freight arriving into sites from the local area • Loading of vehicles for the first/next leg of the freight’s journey to its destination • Sorting of freight arriving from other sites
Throughout the simulation, data is collected relating to various elements of the operation. This allows for in depth analysis of the performance of the operation through graphical and tabular reports. For example, it is possible to understand what freight was left behind at a site and the reason why. This enables focused refinement of the operation to reduce these occurrences.
Calibration of the simulation was critical to ensure accuracy of the results. This involved comparing performance outputs from the NPM with the performance of a reference day in the business. It is then possible to run a range of operational scenarios to identify more efficient and lower cost methods of operation. This ultimately results in rolling out changes into the physical operation to realise those improvements.