MES Pilot at a Manufacturing Enterprise
The plant doesn't know its real cost of production — ERP calculates by standards, foremen enter data from memory. We explain how an MES pilot on one line reveals the real picture and why cost of production is often 20%+ above plan.
A typical manufacturing situation: the general director is confident that the plant operates at 85% OEE and cost of production is within plan. These figures are shown by ERP every month. The problem: these figures often have no connection to reality. Data is entered into ERP by shift foremen at the end of the working day — from memory, by standards, with systematic underreporting of downtime and defects. Not out of malice — the incentive system penalizes deviations from plan.
Why does this happen: ERP at most manufacturing plants is configured as a financial system. It handles accounting well, but doesn’t see the shop floor. Standards aren’t updated for years and don’t account for equipment wear. Downtime is recorded from memory — and systematically underreported. Defects are partially masked as ‘technological losses.’ Between real production and the system lies a human filter that distorts the picture.
Our approach to this challenge: we propose not a large-scale MES project across the entire plant, but a pilot on one critical line — for example, the line that generates 30–40% of revenue. We install sensors on controllers, connect weighing equipment, and begin collecting data automatically: runtime, downtime with causes, actual raw material consumption, output volume, defect rate. No human filter. The first two weeks of the pilot are usually a cold shower: real OEE turns out to be 20–25 percentage points below reported, cost of production 20%+ above plan.
The practical outcome: the shock of truth turns out to be the best driver of change. Management corrects prices, updates standards, changes the incentive system: foremen stop being penalized for downtime and start receiving bonuses for reducing it. Over 3–4 months, OEE grows 10–15 percentage points, downtime drops by a third, defects decrease. And management approves MES scaling to other lines — without a single question about the budget.
Typical Problem
A typical production picture: ERP calculates cost of production by standards, not actuals. Real data — raw material consumption, downtime, defects — is entered by foremen at the end of the shift, from memory, with systematic underreporting. Management suspects the real cost is above plan, but can't prove it. Pricing decisions are made based on figures that don't reflect reality.
Why This Happens
ERP at most manufacturing plants is configured as a financial system — it handles accounting well, but doesn't see the shop floor. Between the real workshop and ERP — a human filter. Three sources of distortion: (1) standards not updated for years and don't account for equipment wear, (2) downtime recorded from memory at shift end — systematically underreported, (3) defects partially written off as 'technological losses.' Foremen distort data not out of malice — the incentive system punishes them for telling the truth.
How We Diagnose It
ERP without MES is like accounting without a cash register. The system calculates how much raw material should be consumed by standard, not how much was actually consumed. Standards age, downtime is underreported, defects are masked. Management makes pricing decisions based on cost of production that may be 20%+ below reality. Our approach — not a large-scale MES project across the entire plant, but a pilot on one critical line. In one month it reveals the real picture and becomes more convincing than any presentation.
The Right Model
MES pilot on one critical line: (1) automatic data collection from line controllers — runtime, downtime, speed, (2) raw material tracking via weighing at input and output, (3) integration with ERP for actual data transfer, (4) OEE dashboard for management and shift supervisors. After the pilot — scaling to other lines.
How We Implement It
We survey the production line, identify data collection points. We install sensors, connect to controllers, develop the MES module for data collection and normalization. We integrate with ERP through an intermediate database. We build an OEE dashboard with shift-level detail and downtime cause breakdown. No human filter between production and the system. A typical pilot takes 3–4 months.
How the Team Works
Projects like this run with a team of 5: 1 developer, 1 process control engineer, 1 integration specialist, 1 business analyst, 1 BI specialist. I define the MES module architecture, ERP integration model, and analytics model. The team implements data collection, integration, dashboards, and testing.
Results
If your ERP shows 85% OEE and cost of production within plan, but data is entered manually by foremen — the real picture is likely different. An MES pilot on one line will reveal the truth in a month. And that truth will be the best driver of change.
Key Lessons
- • If your ERP calculates cost by standards and data is entered manually by foremen — you're making pricing decisions based on fiction. An MES pilot on one line will show the real picture in a month.
- • Start with a pilot on a critical line — when management sees a real 61% OEE instead of the reported 85%, questions about the scaling budget don't arise.
- • Foremen underreport downtime and defects not out of malice — the incentive system punishes them for the truth. Change the incentives in parallel with the MES rollout.
- • MES-ERP integration doesn't always require expensive vendor connectors — an intermediate database with ETL is often sufficient and substantially cheaper.
Similar challenge?
Tell me what's going on. I'll review the situation and suggest a concrete path forward.
Обычно отвечаю в течение нескольких часов