A decision framework to avoid replacing a machine that still has a solid future — or extending equipment that has become structurally risky.
1. Start with the mechanical base, not the PLC
An old machine is not automatically a poor retrofit candidate. The first question is whether its structure, accuracy, guides, loads, rigidity and general condition still match the production objective. An obsolete PLC can be replaced; unstable or worn mechanics can make a retrofit economically weak.
2. Measure the real performance bottleneck
Before writing “automation”, isolate the constraint: operator time, cycle variability, material waiting, quality inspection, manual rework, machine downtime, recurring defects or changeovers. A useful project treats a measurable bottleneck and defines how the gain will be accepted.
3. Assess obsolescence function by function
Obsolescence can mean unavailable spare parts, unsupported software, missing backups, drives with no direct replacement, proprietary networks, incomplete documentation or ageing safety components. A function-by-function matrix is usually more useful than saying “the machine is old”.
4. Include safety and compliance in the decision
A retrofit that changes controls, access, speed, operating modes or guarding can require a serious reassessment of risk and documentation. The question is not only “what does the new PLC cost?” but also “what must be verified, validated and documented after the modification?”.
5. Price the downtime, not only the CAPEX
Two solutions with the same purchase price can have very different economics if one requires three weeks of downtime and the other two days. Scoping should include intervention windows, testing, restart, contingencies and, where needed, a rollback strategy.
6. Check five-year maintainability
The retrofit should leave a machine easier to maintain: clear backups, identifiable components, updated drawings, available spares and an architecture that the maintenance team can understand. Replacing one form of obsolescence with another proprietary dependency is not progress.
7. Use a simple decision matrix
| Criterion | Retrofit more favourable | Replacement more favourable |
|---|---|---|
| Mechanical base | Healthy, accurate, durable | Worn, distorted or limiting |
| Functions to modify | Targeted and identifiable | Deep process transformation |
| Acceptable downtime | Short or schedulable | Long shutdown already accepted |
| Parts / support | Architecture can be modernised | Proprietary system with no path |
| Safety | Upgrade is manageable | Global redesign required |
| Future evolution | Modular additions possible | New architecture required |
Useful output before deciding
A serious pre-study should at least deliver: current-state survey, obsolescence list, functions to rework, safety impacts, principle architecture, downtime assumptions, acceptance criteria and an economic order of magnitude. That is the basis for comparing retrofit and replacement without commercial bias.
General technical scoping note. Regulatory applicability, performance and responsibilities must be confirmed project by project.