Different metaheuristics (e.g., ant-colony-optimisation) have been implemented for dynamic-picking that enable real-time adaptation depending on customer demand and operational requirements. These are based on location/timing and adapt to accommodate incoming orders.
Mathematical models were designed to deal with multiple pickers and tunned to yield new routes rapidly (speed-of-convergence <10-seconds) essential for short timeframes in micro-fulfilment picking scenarios.
Logidot's 3D warehouse simulator is used to evaluate different optimisation methods and train algorithms based on operational data inputs.