Process optimization using predictive analytics by Caduceus
Category: Software Development
Branch: E-Commerce
Company Size: 501 – 1000 Employees
Location: Romania (RO) / Switzerland (CH)
Duration: approx.6 months
Customer request/problem
Our client is an online retailer in the fashion sector. Due to the great success, the company has started to offer many other items, which resulted in very different requirements in the staff scheduling. For example, only a few workers are needed to pick orders from the fashion sector. But when processing orders for furniture items, there is an increased need for staff. Due to the different fluctuating and in many cases unpredictable orders, bottlenecks constantly arose at the individual workstations. Either there were too few workers at one station or too many, who were missing at other places. The warehouse managers reacted to sudden changes in staff requirements by making phone calls. The bottlenecks could rarely be resolved in time, so the average delivery time of the orders increased, which in turn led to a rising cancellation rate and had a negative impact on the company’s performance.
Predictive analytics with our Caduceus solution should solve the problems.
Implementation
By using Caduceus, the entire picking process is monitored. Three KPIs are calculated in real time for each individual workstation: Order status, throughput and the resulting workload. The entire process chain, including all workstations and KPIs, has been visualised on a single dashboard, providing the logistics manager at the headquarters with an overview of the entire process, which spans three locations, for the first time. The manager can now allocate staff efficiently by assigning workers from low workload stations to high workload stations.
By connecting the ordering systems as well as the ERP in addition to the Warehouse Management System (WMS), the order stocks of the respective stations could be projected one hour into the future. As a result, the dashboard shows the bottlenecks in the future and the logistics manager now has one hour to correctly allocate the staff to ensure the efficient execution of the picking process. This reduces the average delivery time and the cancellation rate.
In order to further simplify the work of the responsible logistics and warehouse managers, automatic notifications have been set up in case of workload overruns at individual stations. This eliminates the need to regularly check the processes through the dashboards, as the monitoring system notifies those responsible in good time when their attention is needed and staff needs to be reallocated.
Challenges
In order to implement real-time monitoring using Caduceus, the proprietary interfaces of the individual customer systems had to be connected. The almost exclusive localisation of the interface fields in Romania presented the development team with a particular challenge.
The WMS used by the client was a highly customised standard system and the relevant documentation was not available.
Benefit for the customer
The client was able to solve the challenges in their processes in a sustainable way. Through the Caduceus-based solution, the client now has a tool to continuously optimise all its processes.
Similar Projects
