Automated material handling systems (AMHS) play an increasing role in modern warehouses. They provide speed and efficiency, but if something goes wrong, it can cause big problems. That is why testing these systems is so important. Think of order picking systems that do not reach the promised capacity or a WMS that does not communicate properly with the machines. You can often avoid these kinds of problems by having a well-thought-out testing process. In this article, we explain how to properly test AMHS and what you all need to consider.
What is being tested?
Upon delivery of an AMHS, the supplier often carries out extensive testing together with the customer. The goal? Making sure that everything works as it should. Here are a few examples of tests that are carried out:
- Interface testing: Does the communication between the Warehouse Control System (WCS) and the Warehouse Management System (WMS) work?
- Functional testing: Are the processes within the WCS correct?
- User interface testing: Are workstations user-friendly?
- Capacity testing: Can sorting machines handle enough volume?
- Productivity testing: Do order picking stations achieve the desired speed?
- Quality testing (availability testing): How reliable is the entire system?
Yet in practice, testing is often not done as well as thought. This sometimes leads to unexpected problems after commissioning.
Defining what a system should be able to deliver
Specific performance requirements are set for an automated system, such as the number of order lines to be processed per hour or the number of error-free sorted boxes. These requirements form the basis for the tests and must be fully reflected in the design, environment and results of the test. The system must prove that it can process orders (pick, pack and sort) flawlessly and quickly, while operators can perform their tasks efficiently and for long periods of time. These requirements are translated into detailed tests, where each part of the system is assessed individually. The detailed requirements per level guide the tests and ensure that all aspects are thoroughly validated.
Preparing for testing
Preparing for testing starts by asking exactly what needs to be demonstrated and under what conditions. For example, it is essential that the right materials are present in the order-picking system when testing picked order lines. It is important to include in the test scope both relevant aspects (such as the technical capacity of the system) and exclude unrelated factors (e.g. operator productivity).
The test scope is often determined at different levels, from low (such as correctly connecting control cables) to high (e.g. the operation of planning functions in the control system). This involves looking at function, capacity and quality of the system. The engineering V model is often used to logically compose the tests and determine their scope.
All testing should come together in a clear test plan. This plan should be coordinated with and approved by all stakeholders, such as the project manager, operational users and other stakeholders. Their input is crucial to ensure tests are complete and relevant.
Figure. The engineering V-Model for specification and validation based on system requirements.
Here, it is also important to specify not only the conditions under which a test is successful, but also what should happen in case of an unsuccessful outcome of a test. Defining in advance what the acceptable next steps are, in case of an unsuccessful conclusion of a test, also means that this can be taken into account in a testing process.
Making sure tests validate the right things
Automated warehouse systems come in many forms and functions, but often share similar components. Think mechanical and electrical components, PLC controls, planning systems (warehouse control systems) and integrations with a WMS, for example. All these elements are assessed for functionality, capacity and quality, with specific requirements from customers, engineering projects and users. For example, we check whether user interfaces are user-friendly and workstations are ergonomically designed. External organisations, such as FEM (for quality and capacity) and UN (for GDP), also draw up guidelines for these systems.
When is the best time to test?
Quality tests, such as an availability test, work best under realistic conditions. This means sufficient volume in the system and a test period that is long enough. Often the optimal moment is after go-live, when the system is processing about 50% of the maximum volume. At that time, all conditions are often present: orders, stock, well-trained operators and service engineers with sufficient knowledge. This makes it possible to identify problems that may arise later in the operational phase.
Focus and structure in the test plan
A structured approach to testing based on the system scope and proper requirements are essential for the validation process. Aligning the tests with the requirements and then performing these tests under the right conditions provide the best opportunities to validate the function, capacity and quality as agreed by the supplier and customer. This virtually eliminates the chances of unwanted surprises in the near future. Because in the end, it's all about a smoothly functioning warehouse that everyone can rely on.
Author: Peter te Baerts of St Onge Company, www.stonge.com
Contact: ptebaerts@stonge.com
