Pick Again: Self-Adaptive Warehouse Commissioning

Conference: ARCS 2017 - 30th International Conference on Architecture of Computing Systems
04/03/2017 - 04/06/2017 at Vienna, Austria

Proceedings: ARCS 2017

Pages: 7Language: englishTyp: PDF

Personal VDE Members are entitled to a 10% discount on this title

Meisch, Oliver; Peet, Gerben (Kramp BV)
Rudolph, Stefan; Haehner, Joerg (Universität Ausgburg, Germany)
Mammen, Sebastian von (Universität Würzburg, Germany)

Picking in warehouses represents a central process for mail order businesses. It has a major impact on the required investments by the business and the speed of delivery to the clients. In fact, due to the steady rise in expectations regarding delivery times – even private households receive goods on a same-day-delivery basis, nowadays – ongoing optimisation efforts are rather important. They are all the more necessary, as the economic environment changes rapidly, which includes emerging or receding trading channels as well as the generation of vast amounts of sales and usage data. These dynamics, in turn, equally challenge traditional optimisation approaches and monolithic IT systems. Therefore, in this collaborative project between academia and industry, we have spatially re-modelled an existing small parts store, modelled the workers, dollies, and goods, and optimised various aspects such as the chosen routes and storage locations of the goods. Altogether, we achieved a significant increase in efficiency: The picking process yielded 14% more picks per time and the walking distance was reduced by 37%. Our agent-based, and Q-learning-based approach lends itself well for adapting to changes in the environment as well as changes in the clients’ shopping behaviours.