Kühne + Nagel ranks among the top three worldwide logistics players and provides solutions in the industries Seafreight, Airfreight, Road & Rail Logistics and Contract Logistics. The organization has about 55,000 employees and has network of 900 offices in more than 100 countries.

In close collaboration INCONTROL supported Kühne + Nagel in two different projects of Kühne + Nagel in the food retailing market. The first project concerned the automation of a washing facility for crates. The second project concerned the automation of an order-fulfillment system of customer pallets with the use of "Automated Layer Picking". Both projects are discussed below.

Crate Washing Facility: Introduction
This project of Kühne + Nagel concerns the automation of a washing facility for crates. These crates are used in the food retailing market (e.g. supermarkets) mainly for the transport of fruit & vegetables. For the washing of these crates a machine ("washing street") is used. The crates (6 different types) are internally transported in stacks and on different containers (e.g. rolly, pallet). Several material handling processes (e.g. stacking, palletizing) and transport processes (e.g. robots, conveyors) are needed to support these different types of transport batches. The capacity of the several material handling processes and washing process is dependent on the type of crate that is handled. Since the arrival and delivery intensity varies per moment of the day and per weekday, working buffers are needed to manage the workload on the washing process. Besides the physical aspects about internal transport and material handling also the work buffer management is taken into account in the project performed by INCONTROL.

Purpose and content
For INCONTROL, the purpose of the project was to deliver a user-friendly simulation model. The simulation model contains all relevant settings which enable the user at Kühne + Nagel to analyze and evaluate all relevant considerations of the system design. The available settings concerned the following issues:

• Intensity and content of inbound and outbound flows
• Throughput capacity of the several processes (e.g. crates per hour)
• Use of capacity (e.g. one or two washers, amount of robots)
• Workflow management (e.g. allocation of crates to conveyors)
• Buffer management (e.g. minimal and desired buffer levels that triggered a washing algorithm with corresponding type and batch size)

In order to do a proper analysis of the scenario, the following output parameters were exported to Excel:

• Scenario settings
• Buffer levels
• Decision log (e.g. system state with corresponding triggers)
• Utilization of processes
• Input and output statistic of the system and individual processes (e.g. robots, conveyors)

For commercial and communicational purposes the model was constructed in a 2D and 3D visualization and was based on a CAD drawing.

The use of a simulation model for the washing facility created the following benefits for Kühne + Nagel. They were able to define the appropriate capacity that was needed. So validated decisions could be made concerning capital intensive resources, for instance, the amount of robots, conveyors and washing machines. Also Kühne + Nagel was able to analyze the effect of different growth scenarios which prepared them for future changes in the market. From the initial project that was performed by INCONTROL it was concluded that the designed "washing batch algorithm" did not cope with all possible states of the system. Therefore an iteration was made in this algorithm in order to optimize it. If this optimization had to be done in a later stage of the design or even production, the financial consequences would have been much bigger.

Automated Layer Picking Project - Introduction

The second project concerned a system for automated order fulfillment of customer order pallets in the food retailing sector. In order to perform this order fulfillment, Kühne + Nagel proposed the use of an automated storage retrieval system (ASRS) in combination with an automated layer picker (ALP). The ASRS is used to handle and store both full product pallets and customer pallets. Dependant on physical size and weight of the product, a full product pallet consists of layers that contain (e.g. crates of soda). Since the order quantities of the customer orders do not all exactly match to the quantity one these product pallets, customer pallets are assembled by picking a combination of layers of a product pallet (“mother” pallet) to a customer (“daughter”) pallet. The combination and sequence of these layers is pre-defined by the customer and also on the physical aspects of the products in these layers. In order to manage the workflow of the ASRS with the ALP an optional pre-sorting buffer was proposed to handle the sequencing and working buffer of the ALP.

Unlike the first project, in which also an algorithm for order management was tested, in this project the input of the model is a fixed (deterministic) set of order lines (ALP movements) that are handled by the system. This set of order lines is created with a pre-defined algorithm which functions outside the model. Herewith Kühne + Nagel maintained the flexibility of changing the algorithm without the need of changing the simulation model.

Purpose and contents
INCONTROL was asked to develop a user-friendly simulation model in this project. Like in the first project, the model also contains all relevant settings and is visualized in both 2D and 3D for commercial and communicational reasons. The settings of the simulation model are here:

• Inbound and outbound flow of full product pallets
• Physical properties of the material handling entities (e.g. conveyor speed, handling speed of ALP)
• Use of capacity (e.g. amount of daughter pallet locations, use of pre-sorter, number of ASRS aisles)
• Input of order lines

A feature that is used in this simulation model is the 3D visualization of output monitors that show the key performance indicators. In this way the user can also analyze the system performance in a 3D view. In order to do a detailed analysis of the scenario, the following output parameters were exported to Excel:

• Scenario settings
• Statistics (input, output average content) of the several locations (e.g. ASRS, Pre-sorter, inbound conveyor, ALP daughter positions)

The use of a simulation model for the automated layer picker created the following benefits for Kühne + Nagel. Like in the other project, Kühne + Nagel was able to make validated decisions concerning the capital intensive resources (e.g. ASRS, ALP, pre-sorters). Since the system did not exist yet and the model was mainly focused on 3D visualization, communication concerning the performance and usefulness of the system with both the customer and the supplier was improved significantly. As stated above, pre-defined order lines were used as input of the model, this gave Kühne + Nagel the opportunity to keep optimizing the order algorithm without changing the model.

Conclusion
With the use of the simulation models Kühne + Nagel is able to evaluate important design decisions concerning the capacity and order/buffer management of the systems. Since both systems are not built yet, these decisions can be made early in the design process of the systems. This way validated initial investment decisions about capital resources and order management systems to be made. INCONTROL hopes to continue the support and collaboration with Kühne + Nagel in future projects.