Thanks to our forecasting solution, Mode Logistik GmbH, service and logistics company of Peek & Cloppenburg (Düsseldorf), is able to optimally plan the incoming goods quantities of millions of fashion items and thus required personnel and warehouse utilization.
Challenge
Mode Logistik GmbH & Co. KG is the service and logistics company of Peek & Cloppenburg (Düsseldorf)* Fashion Group Düsseldorf. The customer operates multiple warehouses in which several millions of products of different brands are handled each year. In the fashion industry, delivery dates and goods receipts are difficult to plan. Instead of exact delivery dates, approximate time spans are predominant. The incoming goods quantities vary greatly due to different suppliers as well as the fluctuating delivery reliability of suppliers. Thus, planning of required personnel and warehouse utilization is difficult.
Approach
In order to improve the existing forecasting approaches, all suppliers were first evaluated for their adherence to delivery dates. Based on open orders, the corresponding delivery time windows and the expected delivery accuracy of the suppliers, the expected values of the quantities of the incoming goods down to article level were calculated. These expected values then in turn served as input for a machine learning model, which was extended by external factors such as vacations, public holidays and inventory periods in the warehouse.
Result
Following the outlined analytical approach, i.e. supplier evaluation combined with machine learning forecast, the customer is able to precisely predict the expected delivery date of millions of fashion items from the corresponding suppliers. The success is reflected by an increase in forecast accuracy: the forecast error was reduced by up to 51% for an individual warehouse. The results not only help the customer to better plan personnel and warehouse utilization but prospectively also enable a more intelligent management of suppliers. For example, peaks in capacity utilization could be anticipated and smoothed in advance by distributing orders.
*There are two independent Peek & Cloppenburg companies with their headquarters in Düsseldorf and Hamburg. This reference refers to Peek & Cloppenburg KG, which is based in Düsseldorf and whose locations can be found at www.peek-cloppenburg.de.
THE PROJECT AT A GLANCE
Use case | Prediction of incoming goods quantities (Forecasting) |
---|---|
Industry | Retail & Logistics |
Data basis | Orders, delivery time windows, delivery status, vacations, public holidays, inventory periods in the warehouse |
Project duration | continuously in use after Proof of Concept (approximately 4 weeks) |
Tools | R, Python, MongoDB, PostgreSQL, Angular, Ruby on Rails, Docker |