In a factory, the question 'how many machines do you have' usually gets three different answers. There is one number on the fixed asset list in accounts and another in the maintenance team's logbook; and when you walk the site you find a compressor, two forklifts and a generator that appear on no list at all. The machine's serial number is on its plate but in none of the records; the warranty document is in a folder, and only one person knows which folder. In this picture, none of the questions — where the equipment is today, when it was bought, how many times it has failed and how much has been spent on it to date — can be answered quickly or reliably. Before they can be answered, somebody has to walk the whole site again.
The cost of this usually appears not as one large item but as small, scattered payments. A failure on equipment still under warranty is paid for out of pocket, because nobody is tracking the warranty end date. The same spare part is stocked separately by two different departments, and when it is needed both turn out to be short. A crane or pressure vessel subject to mandatory periodic inspection triggers a hunt for documents the moment it comes up in an audit. The most expensive consequence is replacement decisions: whether a machine should be replaced or not is decided on experience and guesswork, because what has been spent on it over the years is not known. Yet in most plants that decision is the largest spending item of the year.
Enterprise asset management (EAM) keeps machines, plant, vehicles and equipment on a single record from the moment they are bought to the moment they are disposed of. Assets are defined as a tree of site, line, machine and sub-equipment; each is given a unique identity, a label that can be read on site, a technical file, warranty information and a criticality rating. Every maintenance job carried out on it, every spare part used, every stoppage experienced and every invoice paid accumulates on that record. The total cost of ownership of the equipment therefore emerges by itself over time; a decision to replace, to hold a standby or to dispose can for the first time be supported by data, and the argument stops running on guesswork.
The subject of this page is not maintenance work itself. Raising a maintenance order, following the job through and generating a pre-failure warning from the condition of the machine belong to maintenance management and predictive maintenance; asset management, by contrast, manages the whole life of the asset on which that work accumulates. The two can live in the same database, but they answer different questions and their scopes are planned separately. To be honest, the hardest part of the job is not the software either: producing the first inventory on site, labelling it and matching it with the existing records is field work that takes time and manpower. Rather than making that step look shorter than it is, we plan it in stages; we start with critical equipment and do not try to count the entire fleet in one go.