For every machine that is not a CNC.
Presses, furnaces, moulding machines, packing lines and the conventional fleet — three routes in, what each delivers, and the limitations stated plainly.
A great deal of Indian manufacturing capacity runs on equipment that has a PLC rather than a CNC control, or has neither. Presses, moulding machines, furnaces, pretreatment lines, packing machines, special-purpose machines and the entire conventional fleet fall into this group, and they are frequently the machines nobody has ever had a timeline for.
This page describes how monitoring works for that equipment. It is deliberately vendor-neutral: what matters is whether a signal can be reached, not whose logo is on the panel.
Work top-down per machine. Each step adds either configuration effort or an installation slot, so the first route that works is usually the right one. A mixed floor typically ends up using all three, which is expected rather than a compromise.
| Signal read | Typical source | What monitoring derives from it |
|---|---|---|
| Cycle running | PLC tag, or drive contactor | Running, idle and stopped states with exact durations; availability |
| Part or stroke complete | PLC counter, existing counter output, or proximity switch | Output counts, cycle time distribution, performance |
| Mode selector | PLC tag, or selector switch contact | Separates setup from downtime — without it both figures mislead |
| Alarm or fault present | PLC tag, or panel indication | Distinguishes a fault stop from an idle stop before anyone is asked |
| Process values | Modbus register on a controller or meter | Temperature, pressure, flow, current, kWh — trended against machine state |
| Stop reason | Operator at the machine | Why it stopped. No signal supplies this; a person does, in two taps |
For common device classes — energy meters, temperature controllers, drives — MachineWise carries ready-made read profiles, so a device that has been seen before is connected without repeating the register-mapping work. That covers a useful share of the equipment on a typical Indian floor.
For plant-specific PLC logic there is no shortcut, and pretending otherwise is how monitoring projects produce confidently wrong numbers. The tags that represent a cycle and a good part on your special-purpose machine exist only in your machine's program, and mapping them is a conversation between your maintenance team, your production team and ours. The method is set out in PLC signal mapping for machine monitoring.
You get what the PLC knows
If the logic never distinguishes a good part from a cycle, monitoring cannot either until the mapping — or the logic — accounts for it.
No controller-level machining detail
Program names, spindle load and feed override belong to CNC controls. PLC-based equipment does not have them to give.
Modbus carries no meaning
A register is a number until it is mapped, and mapping errors produce plausible values rather than obvious faults. Verification against the device's own display is not optional.
Program changes can break a mapping
If a PLC program is modified and a tag moves, monitoring reads the old address. The mapping document belongs with the PLC backup.
A half-hour survey per machine establishes which of the three routes applies. Machines with a readable PLC are typically connected without stopping production; machines needing wiring take an afternoon each, scheduled around a planned stop. Both are verified by watching machine and dashboard together for a shift before anyone is asked to rely on the data.
A free pilot covers two machines for thirty days. On a PLC-heavy floor we would usually suggest one machine of each type — one readable PLC and one hardwired — because that answers the two questions that determine the shape of a full rollout.