Stacking problems are central to multiple billion-dollar industries. The container shipping industry needs to stack millions of containers every year. In the steel industry the stacking of steel slabs, blooms, and coils needs to be carried out efficiently, affecting the quality of the final product. The immediate availability of data - thanks to the continuing digitalization of industrial production processes - makes the optimization of stacking problems in highly dynamic environments feasible.

In this competition, the goal is to write a policy that optimizes crane operations in simulated environments. The policy receives the current simulation state as input in intervals of 1 second and may send a crane schedule as a response or at any other time that contains moves for the crane, e.g. relocate block 15 from stack 1 to stack 2. The policy must be run the participant's infrastructure (e.g. laptop) and must communicate with our simulation environment in real time. Participants can develop and test their policies against various training settings and score it for a final ranking against unknown scenario settings. The policies are ranked based on a lexicographic objective. Leaderboards are available.

This year, we will host an upgraded version of the crane scheduling (CS) track. It represents a scenario based on real-world warehouse operations, which focuses mainly on the scheduling aspect of crane operations. It features various arrival and handover locations and much more buffer locations to choose from, but a block can only be placed at locations that are compatible with it. Blocks enter and leave the warehouse at arrival and handover locations, respectively, and two cranes can be manipulated to move blocks around. Both cranes share a single lane and can access every location. The scenario creates move requests which determine blocks that have to be picked up at arrival or dropped off at handover locations. An external optimizer must create crane schedules to fulfill these requests. To do so, must create optimized crane moves and respective schedules. Therefore, this scenario integrates three optimization problems: stacking, assignment and scheduling.

The dynamic environment is implemented in form of a real-time simulation which provides the necessary change events. The simulation runs in a separate process and publishes its world state and change events via message queuing (ZeroMQ), and also listens for crane orders. Thus, control algorithms may be implemented as standalone applications using a wide range of programming languages. Exchanged messages are encoded using protocol buffers - again libraries are available for a large range of programming languages. As in the 2025 competition, a website will be used that participants can use to create experiment and test their solvers. In addition, the simulation models are available at GitHub for offline testing and development at https://github.com/dynstack/dynstack . We gladly accept pull requests for new starter kits, existing algorithms and approaches, as well as additions to the bibliography on works that have used the competition for scientific research. Furthermore, we greatly encourage participants to submit a two-page competition entry that explain their optimization approaches.