A sequential approach to truck and internal transport scheduling in a ground handler terminal

A case study

More Info


This study investigates the problem of prioritising trucks at the export area of the KLM Cargo ground handler terminal while maintaining a continuous operation in the unloading area. Due to the uncoordinated nature of truck arrivals and space limitations at the ground handler, substantial waiting times may arise when
trucks arrive simultaneously. This study proposes a sequential approach to solve this problem. First, trucks arrive over time and are scheduled for unloading at doors that can accommodate their cargo type, while some trucks must visit multiple doors. The truck scheduling model utilises a hierarchical objective with the main aim of minimising the number of airway bills missing their outbound flight. Next, the truck’s inbound cargo is grouped into batches and scheduled for transport to their next destination inside the warehouse, while placing an upper bound on the batches queued for transport. The second model aims to minimise the time a batch is queued. This study proposes three mathematical formulations, two for truck scheduling and one for internal transport operations. They are evaluated and compared considering computational time, time to
best, and optimality gap as key performance indicators. Internal transport operations have also been tackled when solving larger instances by utilising a rule-based heuristic to evaluate different prioritisation techniques and reduce computational effort. Finally, a scoring heuristic has also been developed to potentially replace or support the truck scheduling mathematical model, as this is a goal of KLM Cargo to handle trucks in quasi-real time. A parameter tuning of such heuristic has been carried out and its performance against the benchmarking mathematical model has been assessed. It is found that the dynamic scoring heuristic can reduce the number of late airway bills by up to 74 against current operations. Moreover, it is found that by utilising different prioritisation mechanisms, the time batches are queued for transport can be reduced by at least 10 percent.