Digitizing production: choosing the first use case

Which production processes are suitable starting points?

Anyone looking to digitize processes should first check whether a trigger, an intended outcome and the responsible people can be identified for the process in question. The information required must also be established, including whether it already exists or can be reliably recorded during the activity. Material provision, documenting an inspection or handing a batch over to shipping may be suitable candidates. These questions are a selection aid, not a universal suitability test.

A process does not have to run the same way every time. Activities involving expert judgment can also receive digital support if the required information, responsibilities and handling of unresolved questions can be described. Where these points remain unclear, resolving them forms part of the preparation effort. Suitability and priority are two different questions: Only once a meaningful need for support is apparent can the team assess which case to address first.

A production order is marked “complete”, yet the associated goods cannot be handed over to shipping. Quality inspection keeps its results separately. Establishing whether all required inspections are complete and the responsible person has made a decision takes several enquiries.

This is the starting point of our fictional example: order PA-306 concerns the filling of batch CH-082. Order data is held in the planning system, execution is recorded on a paper form, and inspection results are filed separately. The order and the batch have their own identifiers. The company has specified that the batch may only be handed over to shipping once the planned inspections are complete and release has been documented. This is the operational rule for this example, not a general requirement for all production.

Before selecting a solution, it is worth comparing different starting points:

Assess the candidate by frequency, consequences and available data

For manual processes, too, it is useful to consider the activity itself separately from the information work associated with it. For PA-306, a person may continue to supply a container by hand while its association with the order and batch is recorded digitally. The paper form is therefore a reason to look more closely: Does information have to be copied repeatedly? Is an enquiry sent to the wrong person? Is the processing status missing at handover? The number of paper pages saved alone does not describe the operational benefit.

These approaches can be combined. For the initial selection, however, what matters is which one addresses the observed problem. If equipment frequently stops for no apparent reason, recording its states may take priority. In the PA-306 example, by contrast, the missing element is an overview of the prerequisites for handover to shipping. More machine data alone would not answer that question.

For each candidate, those involved should collect three pieces of information: How often does the case occur in a typical week? What consequences does an error have today? And is the required information already available? Frequency includes both the number of affected operations and the number of enquiries actually needed. Consequences may include rework, delays or additional inspection effort; monetary amounts should only be used if they can be established on a substantiated basis.

A rare case with serious consequences may be more important than a frequent minor inconvenience. Conversely, a conspicuous problem does not automatically make a suitable first digitization case if its causes remain unclear. If an inspection result is missing because nobody initiated the inspection, responsibility must first be clarified. An additional input field is not enough.

Weske assigns the investigation of processes and their organizational and technical environment to the “Design and Analysis” phase. Software configuration is a distinct task; the lifecycle allows feedback and does not require a rigid chronological sequence.^[1] As a practical selection aid, this suggests first describing the problem and the required process, then assessing what technical support is missing.

Which information belongs with the production order?

For PA-306, those involved use a specific operation to identify the information needed for handover to shipping. For each item, they record which system or record currently holds it and who can explain its operational meaning. If the use case also covers traceability of the batches used, the finished product identifier alone is insufficient: documented links to the materials actually used are needed.

The connection begins with the order and its batch. It also includes the instructions valid for execution, the materials actually used, the documented work steps and the associated inspection results. The final decision must refer to the correct subject. A release for a different batch does not answer the question about CH-082.

This initial assessment may reveal three different gaps:

• The information exists and is clearly associated. It needs to become accessible to the people involved at the point where they need it.
• The information exists, but its association is unclear. An inspection result might contain a sample identifier whose connection to the batch is known only outside the record. That connection must be clarified.
• The information has not previously been collected. The team must then determine which activity produces it, who records it and how recording can fit into the work process.

This distinction has a considerable effect on scope. Making existing information accessible is a different task from introducing a new inspection or additional recording at the workplace. Selection should therefore not rely on the assumption that the necessary data will somehow turn up later.

The meaning of an item of information must also be correct. A timestamp may indicate completion of an activity or only the later entry of its result. Anyone using it to investigate waiting times needs to know which event it actually measures. The associated process events must therefore make clear whether they document execution or a later data entry.

For CH-082, a new central database is not assumed at this stage. The selection description specifies which connections must be reliably established. If the entire batch execution, including inspections, deviations, and release, is to be brought together in an Electronic Batch Record, the project extends beyond the handover to shipping considered here. This broader scope must be stated explicitly during selection.

Digitizing quality processes in production

The example requires a distinction between three statements: Filling is complete. The planned inspections are complete. The responsible person has made the release decision. These statements concern different activities and may become true at different times.

Digital recording and processing do not yet mean automatic execution. Recording an inspection result electronically and making it available to the responsible person supports the process digitally. A notification sent automatically under defined conditions would be an automated step. In either case, the release decision could remain with a person. Anyone seeking to digitize and automate processes should therefore determine, for each step, what support or automatic execution is actually wanted.

For CH-082, filling is complete, but one planned inspection result is still missing. The useful digitization case is to make this outstanding prerequisite visible together with the affected order and the responsible person. This requires a clear operational definition of which inspection is expected and how its completion can be recognized.

If a result arrives with an unclear sample association, that association must first be clarified. If a result falls outside the criteria defined by the company, the prescribed handling must be initiated. An electronic record does not decide these substantive questions itself. It can bring together the necessary information and processing status so that the responsible person can assess the case.

Selection therefore includes a conversation at the actual workplace: Can digital work instructions provide the person performing the task with the valid instructions where they are needed? Does the required information arise during the activity? Can the person record it there, or would they have to interrupt their work and enter it later from memory?

For the first use case, it is sufficient to define which work and which decision should be connected. The approval workflow for CH-082 must, for example, make clear which inspection results are needed before the decision and who decides on release. For automated compliance evidence, the results and decisions recorded during the activity must remain associated with the correct batch and work step.

Describe the selected case on one page

Selection becomes concrete when production and the functions involved can describe the same operation. For the example, this description can be captured in a short working template. The template is an editorial suggestion; its entries should be adapted to the organization.

A suitable success measure matches the selected problem and has a known baseline. For PA-306, this could be enquiries per order or the effort required to associate documents. A shorter time to the release decision alone would not be sufficient evidence: It remains necessary to check that the cases considered are comparable and that all planned inspections still take place.

This description supports a reasoned decision on whether organizational clarification, better use of existing applications, or additional software is needed first. If a new application is required, it must guide the work steps for PA-306 and embed the necessary controls and evidence in that process. These requirements provide the basis for selecting compliance software. In the pilot, those involved then test whether the same application, acting as workflow software, reliably supports the intended process, including data transfers and exceptions. For PA-306, the objective remains specific: The prerequisites for handover to shipping must be traceably connected for the correct batch.