OEE for container port cranes cannot simply copy the Availability × Performance × Quality formula from indoor manufacturing. Ship-to-Shore (STS) cranes operate in discrete cycles — hoisting, trolleying, gantry travel, lowering the container — completely unlike the continuous flow of a packaging line. Applying the standard OEE framework without adjusting for this discrete-cycle nature produces a misleading Performance number, because a crane cycle's "ideal speed" depends on variables outside the crane's own control — such as how fast yard tractors are supplied.
This is exactly the "Single-Score OEE Pitfall": when Availability, Performance, and Quality are multiplied into one percentage, a crane can have high Raw Quay Crane Productivity but a very low Net Moves Per Hour because it's waiting on yard tractors — a factor entirely outside the crane operator's control. If management only looks at the composite OEE score without separating its components, it's easy to wrongly blame the crane operations team for a problem that actually sits in yard congestion.
A metric set purpose-built for discrete crane operations
Research at Hutchison Korea Terminal (HKT, 14 STS cranes, 2013-2018) splits Availability into two layers: Overall Availability (Ai), subtracting all scheduled preventive-maintenance time, and Availability per Occupied Time (Ao), subtracting only emergency-maintenance time during hours when the crane has active control power. From there, the HKT framework develops Mean Movements Between Failure (MMBF, based on twist-lock counts) and Mean Time to Repair (MTTR) as real reliability measures instead of a generic uptime percentage.
Partene et al. (2026) add four productivity metrics with different structural sensitivities to delay: Net Moves Per Hour (NMPH) measures real operating intensity during net productive time — highly sensitive to operational delay; Gross Moves Per Hour (GMPH) reflects handling intensity under gross conditions; Working Berth Moves Per Hour (WBMPH) normalizes over gross vessel working hours; Raw Quay Crane Productivity (RQCP) isolates pure crane mechanical efficiency. Critically: RQCP and GMPH show high predictive stability (R² above 0.95) because they're structurally driven, while NMPH is far more volatile (R² around 0.90) because it's sensitive to real-time operational shocks like yard congestion.
OEE for container port cranes: why a single score isn't enough
Problem 1 — A composite OEE score hides the real bottleneck
A crane with high RQCP (good mechanical performance) but low NMPH (waiting on yard tractors) produces an averaged composite OEE score — not enough to distinguish "the crane is performing poorly" from "the crane is performing well but stuck behind yard congestion." If the OEE system only reports one percentage, the crane operations team is easily blamed for a problem outside their control.
Problem 2 — Continuous equipment (conveyors) needs a completely different formula than discrete equipment (cranes)
For continuous Bucket-based Excavating, Loading, and Transport (BELT) systems — used for conveyors and bulk-material excavators — OEE is modeled as a product of four factors instead of three: Availability (available time over total calendar time), Utilization (utilization time over available time, capturing idle time), Speed (equivalent operating time over utilization time, capturing speed losses), and Bucket/Capacity Factor (net operating time over equivalent operating time, capturing losses from under-filled buckets or material swell). Applying the discrete-crane three-factor formula as-is to continuous conveyors misses the bucket-factor loss layer — a genuinely significant output-loss source at bulk terminals.
Problem 3 — Sparse data and temporal asymmetry between SCADA and CMMS
Seaport systems commonly lack fully labeled fault data for training predictive models. On top of that, there's a meaningful time gap between the moment SCADA registers a crane as "back to normal running" and the moment a technician administratively closes the work order in the CMMS — causing automated OEE-calculation engines to artificially inflate the downtime window if downtime is calculated from ticket-close time instead of actual machine-restart time.
The operations engineer's lens: reading the full picture, not just one number
An experienced port operations engineer doesn't report OEE to management as a single number — they present all four metrics (RQCP/GMPH/NMPH/WBMPH) side by side, with a clear note on which reflects crane mechanical performance (RQCP, stable, used to assess equipment condition) and which reflects broader system congestion (NMPH, volatile, used to assess yard-operations coordination). Collapsing these two signal types into one score erases exactly the most useful information this metric set provides.
Illustrative scenario: separating crane performance from yard congestion
This is an illustrative scenario for a common type of problem in the industry, not a specific case from any named port: a container terminal tracks crane OEE with a single composite score, and when that score drops during a shift, management defaults to blaming the crane operations team. After splitting the report into RQCP (crane mechanical performance, still stable) and NMPH (net moves per hour, sharply down), the data clearly shows the issue was a drop in yard-tractor supply frequency during that shift — not the crane or its operator. Separating the two signals directs the investigation correctly instead of wasting time inspecting crane mechanics that had no problem.
Reference table: metric — what it measures — predictive stability — operational use
| Metric | What it measures | Predictive stability | Used to assess |
|---|---|---|---|
| RQCP (Raw Quay Crane Productivity) | Pure crane mechanical efficiency | High (R² > 0.95) | Equipment condition, mechanical performance |
| GMPH (Gross Moves Per Hour) | Handling intensity under gross conditions | High (R² > 0.95) | Overall structural productivity |
| NMPH (Net Moves Per Hour) | Net operating intensity, delay-sensitive | Lower (R² ≈ 0.90) | Yard-operations coordination, congestion |
| WBMPH (Working Berth Moves Per Hour) | Normalized over gross vessel working hours | Medium | Cross-vessel berth performance comparison |
| MMBF/MTTR (HKT framework) | Real reliability and maintainability | Grounded in real operational data | Preventive-maintenance planning |
Conclusion
"A single composite OEE score for a port crane isn't wrong — it's just not enough. The real problem always hides inside the components that got averaged away."
Four things worth doing this week if you're running or evaluating an OEE system for seaport cranes/conveyors:
- Check whether the system reports RQCP (crane mechanical performance) and NMPH (net moves, congestion-sensitive) separately, or only a single composite OEE score.
- If running conveyors/continuous equipment, confirm the OEE formula accounts for the bucket/capacity factor, rather than applying the discrete-equipment three-factor formula as-is.
- Cross-check the SCADA "machine back running" timestamp against the CMMS "ticket closed" timestamp — the gap between them directly skews your downtime metric.
- When OEE drops in a specific shift, split RQCP from NMPH before investigating — if RQCP is stable but NMPH dropped, the issue is operational coordination, not the crane itself.