AI and automation
Shift Handover, Work Instructions, Downtime: AI in Production and Maintenance
Shift handover logs, work instructions, downtime analysis and lessons learned: four AI applications for machinery makers, with setup costs and payback.
AI and automation · Series Use Cases Manufacturing, Part 5
8D reports, complaint clustering, audit preparation and inspection instructions with AI: four applications for machinery makers, costed.
Before every audit, the quality department spends weeks pulling together evidence and reports. Between audits, customer complaints are handled one at a time, each on its own, and nobody sees the pattern behind them because nobody has the time to lay twenty cases side by side. On the line, parts are inspected by gut feeling, because the right inspection instruction cannot be found at that moment.
Quality was not one of the options in our survey. In many companies it falls under documentation. That makes sense, since most of the work is about providing evidence, and evidence means documents.
In the reference company with a hundred employees, the four activities in this group account for 42 hours a week. The largest item is a surprise. It is searching for the right inspection instruction at the workstation, not the 8D report, because the search happens every day at every station.
The tool guides you through the eight disciplines, checks for completeness and suggests hypotheses for the root cause. Root cause analysis does not automatically get better, but it no longer gets skipped.
In the model company: 8 hours a week, 40 percent of which can be automated, €12,240 a year. Setup €4,000, running costs €180 a month, payback after 4.8 months. Implementation: licence and training, no infrastructure of your own, 2 to 4 weeks. Tools: Claude, Babtec, Microsoft 365 Copilot, FORGE Insights (our own product). Data protection: not critical, internal data with no personal data involved.
Cases are grouped by failure pattern and assembly, and the most expensive recurring causes become visible. The time saved is small. The real lever is the repeat defect that never happens, and that does not appear in any calculation of hours.
In the model company: 6 hours a week, 35 percent of which can be automated, €8,032 a year. Setup €3,500, running costs €255 a month, payback after 8.4 months. Implementation: builds on the shared knowledge platform, 3 to 6 weeks. Tools: Claude, Babtec, Azure AI, FORGE Insights (our own product). Data protection: needs review, settle the data processing agreement and deletion policy in advance.
Compiling evidence, generating draft reports, listing gaps in advance. Weeks become days, and the gaps show up before the audit instead of during it.
In the model company: 8 hours a week, 40 percent of which can be automated, €12,240 a year. Setup €4,000, running costs €135 a month, payback after 4.5 months. Implementation: licence and training, no infrastructure of your own, 2 to 4 weeks. Tools: Microsoft 365 Copilot, Babtec, Claude, ComplAInow (our own product). Data protection: not critical, internal data with no personal data involved.
At the terminal, an assistant answers the question of how this part is to be inspected, in the operator’s own language. Access is through shared terminals, so hardly any named licences are needed. It has the shortest payback in the group once the knowledge platform is in place.
In the model company: 20 hours a week, 30 percent of which can be automated, €17,550 a year. Setup €3,500, running costs €255 a month, payback after 2.9 months. Implementation: builds on the shared knowledge platform, 3 to 6 weeks. Tools: Cognigy, Azure AI Search, Nextcloud Assistant. Data protection: not critical, internal data with no personal data involved.
Audit and 8D first: both are pure licence applications with no infrastructure, paid back in under five months. The inspection instruction has the shortest payback in the group and makes a difference at every station, but it builds on the knowledge platform. If you already have one, start here. Clustering customer complaints is the most important application strategically and the weakest in the numbers. It pays off if you measure it by the one defect that does not happen a third time, rather than by processing time.
| Application | Hours per week | Savings per year | Setup | Payback |
|---|---|---|---|---|
| Inspection instructions at the workstation | 20 | €17,550 | €3,500 | 2.9 months |
| Audit preparation and QM reports | 8 | €12,240 | €4,000 | 4.5 months |
| Structuring 8D reports | 8 | €12,240 | €4,000 | 4.8 months |
| Clustering customer complaints | 6 | €8,032 | €3,500 | 8.4 months |
All figures come from the calculation model of the use case library, not from the survey. Reference company with 100 employees, 45 productive weeks a year, fully loaded costs of €65 an hour for commercial staff and €85 for technical staff, licences at €45 per user per month. The model counts only working time saved against the full costs, and the savings rates are deliberately set at the lower end of what is known from projects and studies. No quality gains, no avoided cost of errors, no effects on lead time. Replace the assumptions with your own figures and the model calculation becomes your calculation.
The full calculation for all thirty applications is in the Use Case Library for Machinery Manufacturing, and the managing directors’ answers are in the survey results (in German).
Where is your biggest lever? Let’s talk it through briefly, no obligation.
Read on
AI and automation
Shift handover logs, work instructions, downtime analysis and lessons learned: four AI applications for machinery makers, with setup costs and payback.
AI and automation
Five AI applications around enquiries and quotes in machinery manufacturing, with hours, setup costs and payback for a reference company of 100 staff.
AI and automation
Five AI applications for order processing and purchasing in machinery manufacturing, costed: hours, setup, payback, and where automation does not pay.