Even the best technology will not fix business processes with basic design flaws. The same is true for MES and supported production processes. Digitizing an overly complex or patchy manufacturing process may not yield the benefits you were expecting, e.g., higher efficiency, improved compliance, or better traceability. What is worse, most of today’s software comes with high flexibility and customization. Tempting as it may sound, to just go along with a vendor’s off-the-shelf solution will not necessarily help you define better processes. There are simply too many options, and you may get lost if you short-cut this utterly important step: getting your business processes straight.
The most successful companies start with a clear picture of their business requirements and let these drive their technology initiatives. It begins with understanding your manufacturing style and level of complexity (continuous/batch/discrete), followed by a detailed review of use cases and desired functionality. These steps naturally lead into a MES selection process, guided by a clear scope of the previously-defined data and business requirements. A clear scope is especially important because a system usually shines in the field it was designed for, but needs high effort to be adapted to processes it was never intended for. So optimizing and defining your processes before selecting a solution is essential. In terms of maintenance and development, one must consider that effort increases exponentially the further away from the initial off-the-shelf solution the requirement is.
Here are five perspectives on your business process design that will help you prepare your operations for digitalization:
1. Clarity of business procedure
Since any IT software in the end is a system based on logic, an IT- supported process can only be as good as the business rules defining it. Make sure before designing the process in the system it is well designed (meaning it always follows logical rules) in the real world.
2. No redundancy
A well-designed end to end process should come without any redundancies in terms of data capture (one time only) and documentation. Defining which system has the lead for certain information is important to achieve this. For example: typically the ERP has the lead on product master data, inventory and shipping. An MES has the lead on how a product was produced in detail and on what the shop floor operator in production should do.
3. One role – one system
While a complex end-to-end process might require support of different IT systems it should be always considered that one business role works in as few systems as possible. The goal should be to manage the systems in a way where a normal user only works in one system to do his daily job. There are two main reasons for this approach. Each manual lookup of data, or the management of data independently, in different systems, is error prone and frustrating for a user to constantly switch systems.
4. Minimize handovers
Clearly, in many cases, work in one system cannot be done without data from another system. A process can be vastly improved if the links and handovers from one business area to another are reduced and are clearly defined. Each handover results in an interface requirement with possibilities for errors. Interfaces always should be as few as possible and simplified to achieve a stable and sustainable process. But, as we are in a modern full digital environment, we should not be afraid of interfaces in general. Modern systems often come with supporting tools that minimize the effort of interface developments. The key goal should be to limit the connections while leveraging the core functionalities of the individual systems as much as possible.
5. Keep it simple
Consider excluding exceptional processes from interface automation that rarely occur and instead define a partly manual procedure for such interventions. It may drastically reduce complexity, training requirements and headaches later. Sometimes a well-defined manual process is better than an overdesigned automated IT process.