The application of the 5S methodology to warehouse management represents an important
step for all manufacturing companies, especially for managing products that consist of
a large number of components. Moreover, from a lean production point of view, inventory
management requires a reduction in inventory wastes in terms of costs, quantities and time
of non-added value tasks. Moving towards an Industry 4.0 environment, a deeper understanding
of data provided by production processes and supply chain operations is needed:
the application of Data Mining techniques can provide valuable support in such an objective.
In this context, a procedure aiming at reducing the number and the duration of picking
processes in an Automated Storage and Retrieval System. Association Rule Mining is applied
for reducing time wasted during the storage and retrieval activities of components
and finished products, pursuing the space and material management philosophy expressed
by the 5S methodology. The first step of the proposed procedure requires the evaluation
of the picking frequency for each component. Historical data are analyzed to extract the
association rules describing the sets of components frequently belonging to the same order.
Then, the allocation of items in the Automated Storage and Retrieval System is performed
considering (a) the association degree, i.e., the confidence of the rule, between the components
under analysis and (b) the spatial availability. The main contribution of this work is
the development of a versatile procedure for eliminating time waste in the picking processes
from an AS/RS. A real-life example of a manufacturing company is also presented to explain
the proposed procedure, as well as further research development worthy of investigation.