Environmental & Energy Management System Suite
Machine-Level Energy & OEE Analytics
Most plants run two sets of figures. One describes equipment effectiveness, the other describes energy. The first does not know what the machine drew; the second rarely goes finer than the workshop or the distribution board. So a machine that looks ineffective and a machine that looks expensive are never the same machine on paper.
Machine-Level Energy & OEE Analytics puts both on the same machine over the same period: how much of the shift was producing, what the rest of it was doing, what each of those states drew, and what that works out to per good part produced.
The distinction that matters is that running is not producing. Changeover, idle running and standby all draw power while producing nothing. Keep them apart and it becomes possible to say where that part of the bill went.
Running is not producing.
What it is
One record per machine. Consumption and effectiveness are registered against the equipment itself, not averaged across a workshop or a distribution board.
OEE and its three components. Availability, performance and quality are calculated separately, and losses are broken down to a level someone can act on.
Energy split by what the machine was doing. Producing, changeover, idle running and standby are metered apart, with no-load and standby power identified on their own.
Energy per good unit. The denominator is good parts, not parts made, so quality loss and energy sit in the same table.
Machines compared with each other. Same model, same operation, ranked — so which one to start with is a decision with evidence behind it.
Two-way with MES. Orders, output, good counts and shifts come from MES; OEE and energy results go back to it.
What gets in the way today
OEE and energy are counted separately. Two sets of figures that never meet.
The effectiveness system does not know what the machine drew. And the energy system does not know whether the machine was producing.
Energy data rarely goes finer than the workshop or the distribution board. So it cannot be attributed to a single machine.
Low effectiveness and high consumption therefore never appear on the same line. Which is why the two are discussed as separate problems.
The two systems use different time grains and different rules. So even when both are available, they cannot be laid side by side.
States are guesses. What the machine was doing is assumed, not recorded.
Producing, changeover, idle, standby and down are entered by hand or judged from experience. Which means they are entered when somebody has time.
Night shifts and unmanned hours go unrecorded. The state column is blank, or everything is marked as running.
Running while producing nothing is not noticed. Because nobody is watching continuously, and it looks normal from the outside.
Downtime reason codes are not maintained. Anything can be entered, so nothing can be analysed.
The machine's own consumption is a black box. Nobody knows which unit is expensive.
Only the workshop total is known. Which machine is the large consumer is a matter of opinion.
Two machines of the same model have never been compared. So a difference between them is invisible.
New equipment is selected on nameplate rating. Because there is no measured figure to select on.
Standby and no-load power have never been measured. So whether it is worth switching off is unanswerable.
Nothing to start an improvement from. No baseline, so no target.
How long changeover actually takes is not recorded. So a target for it can only be invented.
Whether standby should be switched off is unknown. Along with what switching it off would affect.
Before and after cannot be compared on the same basis. So the result of a change stays a claim.
Machines of the same type are not ranked. So which to start with is decided by impression.
What it measures on each machine
Five states, and the question of which of them produce and which only spend.
State | Producing | Drawing power | Why it is worth separating |
Producing | Yes | Yes | The denominator and the reference point |
Changeover | No | Yes | How long each one takes and what it draws is directly actionable |
Idle running | No | Yes | Turning without output, and the easiest state to mistake for normal |
Standby | No | Yes, usually less | Small at any moment, substantial across a shift, often with windows where it could be off |
Off | No | No | Kept apart from standby, otherwise downtime is understated |
How state is determined depends on what the site has. Where machine signals exist, state comes from signals. Where they do not, it is inferred from power behaviour or from rules, and anything inferred can be corrected by hand. Ideal cycle time, rated power and state thresholds are parameters configured per machine — they are not decided by the software on the plant's behalf. What this application does is separate the states, calculate the figures and show the differences. Whether a machine on standby gets switched off, and whether schedules change, is decided on site.
What you get
Where it is used
What changes between these settings is which states cost the most, and which losses are worth attacking first.
Setting | What the analysis usually focuses on |
Discrete manufacturing — automotive parts, machinery | Machining centres and line equipment: idle running, changeover duration, comparison between units of the same model |
Injection moulding and forming | Holding power at standby, empty cycles, energy per shot |
Electronics and semiconductor | Process equipment on standby, temperature and humidity loads, losses at line changeover |
Food, beverage and central kitchens | Cleaning and changeover, refrigeration start and stop patterns, standby windows |
High-consumption individual equipment | Compressors, pumps and fans, melting and kilns: load factor and idle draw |
Multi-site operations | The same model compared across plants, separating a difference in configuration from a difference in practice |
Capabilities
Grouped by what they do. Energy measurement at site, area and tariff level sits with Energy Consumption Management; this application works at the level of the individual machine.
Capability | What it means |
Machine register | Record model, rated power, ideal cycle time, line and operation for each unit. |
State detection | Determine state from machine signals where they exist, or from power behaviour and rules where they do not. |
Configurable state definitions | Set what counts as producing, changeover, idle running, standby and off, per machine type and per unit. |
Manual correction | Correct an inferred state by hand, with the correction recorded. |
Output and good count | Take produced and good quantities from MES, from counters, or by confirmed entry. |
Downtime reasons | Maintain reason codes and capture them, so downtime can be analysed rather than just totalled. |
Availability | Time the machine was available against the time it was planned to run. |
Performance | Output against the ideal cycle time for the time it ran. |
Quality | Good parts against parts produced, with scrap and rework counted as loss. |
Loss breakdown | Break the gap down to the losses that can be worked on, rather than a single effectiveness figure. |
Ideal cycle configuration | Set ideal cycle time and standard rate per product and per machine. |
By shift, order and product | Read the same figures by shift, by order and by product. |
Energy per machine | Consumption recorded for the individual unit over any period. |
Energy by state | What producing, changeover, idle running and standby each drew. |
No-load and standby power | Identify what the machine draws while turning without output, and while waiting. |
Start-up peaks | Show the draw when equipment starts, which is often where a large part of the cost sits. |
Energy per good unit | Consumption divided by good parts, so quality and energy are consistent with each other. |
Ranking across machines | Rank units of the same model and the same operation against each other. |
Trends and drift | Show slow deterioration over time, which is what a single period rarely reveals. |
Deviation alerts | Flag a machine drifting away from its own pattern or away from its peers. |
Standby windows | Identify long periods at low power that could be off, and leave the decision to the plant. |
Actions and re-measurement | Raise improvement actions, then measure the same machine on the same basis afterwards. |
Two-way MES integration | Take orders, output, good counts and shifts from MES; return OEE and energy results to it. |
Equipment and automation | Take state and counts from PLC, SCADA and gateways already installed, over the protocols the site uses. |
Deployment choice | Run in the cloud or on your own servers, usually decided by where operational data may be processed and stored. |
Access and retention | Role-based access per site and area, retention set by policy, with access and changes logged. |
How it works
Register. Record each unit: model, rated power, ideal cycle time, line and operation.
Connect. Consumption from meters or gateways, state from machine signals or power behaviour, output and good counts from MES.
Define states. Set state definitions and thresholds, starting from a template per machine type and adjusting per unit.
Calculate. The three OEE components, energy split by state, and energy per good unit.
Compare. Rank units of the same model, watch trends and drift, and identify standby windows.
Act and re-measure. Raise improvement actions, then measure the same machine on the same basis afterwards to see whether it changed.
Integration, deployment and data boundaries
Two-way with MES. Orders, output, good counts, shifts and state times come from MES, and OEE and energy results are returned to it, so effectiveness and energy appear in the same production view.
Equipment and automation. State and counts come from the PLC, SCADA and gateway layers already installed, over whatever protocols the site runs.
Metering. Consumption comes from the meters and gateways already in place. Whether to add more is a decision the plant makes once the data shows whether it is worth it.
How it divides the work with energy management. Energy Consumption Management works at the level of media, areas, orders and tariffs. This application works at the level of the individual machine and its OEE. Both share the same collection rather than building separate accounts.
What is local. Retention periods, access to equipment data and any sector-specific rules depend on the jurisdiction and the customer's own policy. Configuration is set to fit them; the system does not make those decisions.