A machine runs its cycle as long as everything works. Then a sensor fails or the controller needs replacing, and the wiring no longer matches the drawings. With the original PLC project unavailable, the engineer has to work out how the machine moves, what feedback the controller receives and what conditions trigger the next command.
The scope depends on the job. Current wiring diagrams and an I/O map may be enough for repair. Replacing the controller also requires a description of operating modes, transitions and responses to faults.
Start with the process
Follow a normal cycle from the operator’s actions to the finished part. Which mechanisms move, in what order, and what confirms the end of each operation? Record manual and automatic modes, start conditions and how operators usually recover after a stop.
Compare these observations with HMI messages, fault logs and available documentation. Operators can point to recurring faults and help identify where to look for evidence. Use available diagnostics or trials under an agreed test plan to confirm transition conditions.
Reconstruct the wiring and I/O map
Document power supplies, PLC modules, terminals and interposing devices. Trace each command to the actuator and each feedback signal to the controller input. For example, a PLC output energizes an interposing relay, the relay powers the hydraulic valve coil, and a position sensor signals the cylinder’s position to a PLC input.
For each signal, record the PLC address, terminal, purpose, electrical type and active state. Note how you verified it. Here is a sample from a hypothetical cutting machine.
| Signal | Type | Purpose |
|---|---|---|
| DI1 | Input | Cycle start command |
| DI2 | Input | Clamp in working position |
| DI3 | Input | Blade at home |
| DO1 | Output | Lower clamp command |
| DO2 | Output | Blade cutting stroke command |
Verify the connections against the installed wiring. For work that requires shutdown, follow the site’s procedure for isolating energy sources and preventing unexpected startup. Account for residual pressure, electrical charge and mechanisms that can move under their own weight.
Reconstruct the automatic cycle transitions
For the same machine, the sequence can be described as follows.
Idle → clamp → dwell → cut → return → idle.
For each step, record its entry conditions, active commands, expected feedback and timeout response. If the blade starts moving after the clamp sensor changes state, record the timing as an observation. Confirming that the sensor enables the transition requires further evidence from the logic or a test under an agreed plan.
If there is a delay between start and movement, compare the timing of the start command, feedback from the previous step, the PLC output and the actual onset of movement. A delay before the output may come from waiting for a permissive or a programmed dwell. A delay after the command may arise in the relay, valve or hydraulic system. The timing narrows down where to look for the cause.
Examine interlocks and faults
Start with scenarios that affect startup, stopping and recovery. What happens if the clamp position is never confirmed? How does the machine respond to communication loss or restored power? For each scenario, record the source of evidence and mark the behavior as confirmed, inferred or unknown.
A process interlock controls the order in which operations may take place. A safety function reduces the risk of harm and requires its own assessment and verification. The principles in ISO 12100 can guide risk assessment during a retrofit.
Begin with fault logs, built-in diagnostics and records of previous stops. Plan fault simulation on a test bench or under specially prepared conditions using an agreed test plan.
Document the machine and verify changes
For maintenance, collect current wiring diagrams, the I/O map, operating modes, transitions and recovery procedures. Include available backups and test results. For a retrofit, document the requirements for the new controls separately from the description of the existing machine. If a fault response changes, record the reason and how it will be verified.
When replacing the PLC, test the logic on a bench first, then verify inputs, outputs, operating modes and stop behavior in stages on the machine. The documentation should let the next engineer follow a signal from the terminal to the transition condition and see how that condition was verified.