What is the difference between a CMMS and what you build?
A commercial CMMS is a packaged product, and if one fits your operation you should probably buy it. Custom work makes sense when you need integration your plant systems cannot support off the shelf, when your asset or process structure does not fit a product's assumptions, or when a package would require you to change working practices that exist for good reasons. We will say when a package looks like the better choice, and we have done so.
Will our technicians actually use it?
Only if logging a job takes under a minute on a phone, works with no signal, and requires no typing where a photograph or a scan will do. That is the design constraint we start from. Where maintenance systems fail it is almost always here, and the failure looks like a full asset register with almost no work history, which makes every analysis module worthless.
Is predictive maintenance worth it for us?
For some assets, possibly, and we will tell you which. Real failure prediction needs dense sensor data and enough recorded failures to learn from, and most plants have neither across most equipment. Condition based triggers on measured runtime or a threshold are simpler, reliable, and capture much of the benefit. We would rather deliver that than a prediction model whose output nobody can trust.
How deep should the asset hierarchy go?
To the level at which you make decisions, usually the replaceable component on critical assets and the machine on the rest. Too shallow and you cannot see that one component fails repeatedly. Too deep and technicians face a long search at the worst possible moment and start selecting whatever is quickest, which corrupts the data. It is a judgement made with the people who will use it.
Can it work without machine integration?
Yes. Calendar based schedules, work orders, parts, and cost history all function on manual entry, and that is where most plants start. Machine integration improves it by triggering on actual runtime instead of assumed runtime and by raising requests automatically from faults. It is an enhancement rather than a prerequisite, and we sequence it after the basics are working.
How long before the reporting becomes useful?
Cost and downtime reporting needs history, so expect several months before trends mean anything and a year or more before a capital case is well supported. This is worth being honest about at the outset, because a client expecting reliability insight in month two will be disappointed by a system that is working correctly. Immediate benefits come from schedule compliance, spares visibility, and work order control.
What about spare parts we do not know we need?
The criticality analysis surfaces them. Working through which assets stop production and which parts have long lead times routinely finds critical spares nobody holds, and stock held for equipment removed years ago. That exercise often returns more in avoided downtime and released stores value than the software does in its first year.
Can you handle permits to work and safety procedures?
Yes, tied to work orders so a job cannot be recorded as started without the required authorisation, with isolation and safety steps recorded as part of the job. The specific procedures and their legal requirements are defined by your safety function, not by us. We implement them as enforceable steps and retain the evidence that they were followed.
How does this relate to production planning?
Maintenance needs machine time and production needs output, so the two must share a calendar rather than compete for it. Planned maintenance windows block capacity in the schedule, and production priorities inform when work can realistically happen. Our production planning page covers the scheduling side, including how maintenance windows enter the capacity picture.
How long does implementation take?
A focused system with asset register, preventive schedules, work orders, and parts is typically a few months, with hierarchy design and technician interface testing taking a meaningful share. Machine integration, condition monitoring, and multi site standardisation extend it. The organisational work of agreeing a failure taxonomy across sites is often slower than the software.
What do you need from us to start?
Your current asset list however messy, existing preventive schedules, spare parts stock data, a sample of recent maintenance history, an equipment list with controller types, and access to technicians and the maintenance manager. Time in the plant with technicians is the most valuable input, because the interface design that determines success cannot be specified from a meeting room.
