From Group to Shift, six-level cascade, drill-down layer by layer
Based on the Leansight Smart Operations Control Tower system, building OEE cascade management from the Group strategy level to the frontline shift level.
Data aggregates bottom-up, decisions cascade top-down, and OEE at each level is both an outcome and an action.
One formula, three dimensions, six-level cascade
The world-class OEE standard is 85%+, while most manufacturing companies' OEE ranges between 40%–60%.
The value of OEE lies not in "calculating a number," but in "knowing why that number is what it is, and at which level to improve it"—this is the purpose of the six-level cascade.
From Group strategy to shift execution, each level of OEE focuses on different dimensions and serves different decisions
Each level of OEE answers different questions and serves different decision-makers
Answers "Which site has the best capacity efficiency? Who should resources be allocated to?" Group OEE aggregates multiple factory OEEs, benchmarking by region/product line/customer dimension, identifying capacity bottleneck sites, and guiding capital expenditure and production line investment directions. It's not about looking at individual OEE data points, but about trends and benchmarking gaps.
Answers "Which workshop is dragging performance? What's the focus for improvement this month?" Factory OEE aggregates workshop OEEs, benchmarking and ranking workshops, identifying underperforming workshops, and setting monthly improvement priorities. At the monthly operations review meeting, the OEE benchmarking table is the core agenda item, and improvement project initiation is directly linked to OEE loss breakdown.
Answers "Which line is the bottleneck? Which line to improve this week?" Workshop OEE aggregates line OEEs, benchmarking and ranking lines, identifying bottleneck lines and lines with the highest variability. At the weekly workshop improvement meeting, bottleneck line OEE loss breakdown is the core discussion item, producing improvement work orders.
Answers "What's this line's OEE right now? Why is it low? Which shift dragged it down?" Line OEE is broken down by shift, with real-time monitoring of downtime events, speed losses, and defective products. Andon system pushes anomalies in real time, and shift OEE benchmarking drives healthy competition. Downtime root cause analysis drills down from shift level to equipment level.
Answers "What's this shift's OEE? Which shift dragged it down? Where are the losses?" Line OEE is broken down by shift, with real-time recording of root causes for each downtime event, every speed drop, and every defective product. Shift handover dashboard is auto-generated; shift handover no longer relies on verbal communication but on data. Andon system pushes anomalies in real time with second-level response. Shift OEE benchmarking drives healthy competition.
Answers "Which equipment is the bottleneck? When should maintenance be scheduled? What's the root cause of failures?" Equipment OEE is the finest granularity, drilling down from shift OEE to individual equipment's availability (MTBF/MTTR), performance (cycle time deviation), and quality (equipment-related defects). Predictive maintenance provides early warnings based on equipment OEE trend degradation models, and spare parts inventory is linked with OEE trends. This is the endpoint of the six-level cascade—all OEE losses can ultimately be attributed to specific equipment.
OEE is not a static report; it's a bidirectionally flowing operational nervous system
At each level of aggregation, loss structure surfaces simultaneously—what upper levels see is not just an OEE number, but a loss waterfall broken down to A/P/Q.
When targets cascade top-down, each level has clear improvement metrics and action items—not just empty slogans of "improve OEE," but "reduce changeover time of equipment E-07 from 45min to 20min."
The OPS pyramid tells you "who should look at what"; the OEE cascade tells you "how deep the data can drill down"
| OPS Lean Digital Operations Pyramid | OEE Full-Chain Cascade | |||||
|---|---|---|---|---|---|---|
| Level | Role | What They Focus On | Level | OEE Granularity | Driven Decisions | |
| L5External Display | Customer visits Government officials |
Brand credibility Overall operations overview |
→ | L1Group OEE | Group-wide OEE trends | Brand display External visits |
| L4Business Decisions | President General Manager |
Strategic direction Investment ROI |
→ | L1-L2Group/Factory | Cross-factory OEE benchmarking Improvement priorities |
Investment/Expansion Improvement resource allocation |
| L3Business Domains | Directors (Q/C/D/S/M) |
Cross-department collaboration Domain metric benchmarking |
→ | L2-L3Factory/Workshop | Breakdown by domain OEE loss structure |
Improvement project initiation Cross-department initiatives |
| L2Functional Domains | Workshop leaders Department heads |
Cross-process benchmarking Bottleneck localization |
→ | L3-L4Workshop/Line | Cross-line OEE ranking Bottleneck line list |
Improvement order scheduling Line optimization |
| L1Refined Management | Managers Team leaders |
Real-time monitoring Instant anomaly response |
→ | L4-L6Line/Shift/Equipment | Real-time OEE monitoring Equipment-level root cause |
Andon response Maintenance/Shift handover |
The OPS pyramid has more people at lower levels—from the GM to team leaders, management breadth increases;
The OEE cascade has finer granularity at lower levels—from Group OEE to Equipment OEE, data precision increases.
Every mapping point formed by their intersection is a complete closed loop of "who, with what granularity of data, makes what decision."
From theoretical capacity to actual output, every drop of loss is accounted for
The loss waterfall chart gives every drop of loss an attribution—not a vague statement of "OEE is low," but "fault downtime accounts for 12%, of which E-07 accounts for 5.8%"
Six capabilities spanning the complete closed loop of OEE from collection to insight to action
Unified integration of multi-source data from EAP/SCADA/PLC/MES/ERP, with second-level collection of equipment status, output, defects, and downtime events. IoT gateways connect directly to production line equipment.
Stream computing engine calculates A/P/Q rates in real time, with shift OEE refreshing at second-level. Downtime events are automatically categorized (planned/fault/material wait/changeover) without manual entry.
LeanBI dashboards cascade six levels from Group to Shift; clicking any OEE number drills down to the next level's loss breakdown, down to specific equipment, specific shift, specific event.
LeanFusion AI automatically analyzes downtime root causes, correlating equipment parameters, material batches, and personnel shifts, providing structured insights like "Equipment E-07 frequently stops due to temperature drift in XX chamber."
OEE anomalies are pushed in real time to line supervisors' phones, with the andon system simultaneously triggering on-site audio-visual alarms. Tiered push by anomaly severity prevents alert storms.
OEE losses automatically generate improvement work orders, linked to LeanCodee low-code workflows. Improvement results are trackable, benchmarkable, and reviewable. Forming a "discover → improve → verify" closed loop.
Small steps iteration, layer-by-layer cascade, validate first then scale
Select 1 benchmark line, integrate equipment data, achieve three-level cascade: Line OEE + Shift OEE + Equipment OEE. Validate data collection accuracy and OEE calculation logic, establish baseline.
Replicate the pilot line model to all lines in the workshop, achieving four-level cascade: Workshop OEE + Line OEE. Line benchmarking and ranking, identify bottleneck lines, launch weekly improvement meeting mechanism.
Expand to all workshops across the factory, achieving six-level OEE full-chain integration. Monthly operations review meetings use the OEE benchmarking table as the core agenda, with improvement project initiation linked to OEE loss breakdown.
Multi-site OEE benchmarking, viewing trends and gaps at the Group level. Introduce LeanFusion AI capabilities for automatic root cause analysis, predictive maintenance, and OEE trend early warning—upgrading from "looking at data" to "data finding people."
Using a factory with 1 billion annual output as an example, the quantified value of OEE improving from 60% to 70%
The true value of OEE improvement lies not in "more output", but in "less waste"—less downtime, less rework, less overtime, less inventory.
The six-level cascade gives every drop of loss an attribution, making every improvement quantifiable.
Customer M's 8-year digital practice, 50→2000+ dashboards, real-world implementation of six-level OEE cascade
7 iron rules + 5 red lines, the 8-year methodology from IT manager being severely criticized to becoming factory GM
From sowing to sublimation, tree growth metaphor demonstrating the construction of the control tower's four-layer architecture
Autonomous discovery layer—from OEE data to business insights, AI proactively discovers blind spots
The data foundation of the OEE full chain—6 AI capabilities supporting from collection to insight