Data4Sim

Data-driven simulation of manual logistics processes

The project develops methods for capturing manual operating processes with body-worn sensors and transferring them, partly automatically, into simulation models. These simulation models make it possible to predict the effects of changes in manual operating processes, for example in manual picking processes in logistics centers or retail distribution warehouses.

Changes in work sequences, such as variations in pick-list generation, or changes in storage strategies, such as item placement in the warehouse, can be evaluated through simulation before they are implemented.

Technically, the project first records manual processes with wearable sensors. Based on this sensor data, AI methods such as convolutional neural networks infer the activities performed by employees. These activities are transferred into a detailed simulation model of the manual processes. The model represents parameters such as task duration, dependencies such as fatigue or experience, and the temporal structure of activities. The resulting simulation model supports targeted planning and evaluation of process optimization measures.

Publications

Moh’d Khier Al Kfari, Stefan Lüdtke. Domain Adaptation in Human Activity Recognition through Self-Training. Companion of the 2024 on ACM International Joint Conference on Pervasive and Ubiquitous Computing (Ubicomp Workshops) 2024. [web]

Friedrich Niemann, Fernando Moya Rueda, Moh’d Khier Al Kfari, Nilah Ravi Nair, Stefan Lüdtke and Alice Kirchheim. Towards Standardized Dataset Creation for Human Activity Recognition: Framework, Taxonomy, Checklist, and Best Practices. 9th International Workshop on Annotation of Real World Data for Artificial Intelligent Systems (2025).

Contact

Moh’d Khier Al Kfari

mohd.kfari@uni-rostock.de