Definition
Measures the ratio of actual value-adding processing time to total plant flow time for a production unit or batch under AI orchestration.
The Measurement Gap
Traditional production scheduling balances competing priorities through human judgment. A scheduler knows that maximizing utilization at the bottleneck station matters more than maximizing utilization everywhere. They accept idle time at non-bottleneck stations to keep the system flowing.
AI orchestration systems often optimize each station or cell independently, or optimize for a single objective (throughput, utilization, energy) without fully accounting for the systemic flow consequences. A station that runs at 95% utilization looks excellent in isolation, but if its output sits in a WIP buffer for four hours waiting for the next station, the system-level efficiency is poor. The AI made the station look better while making the plant perform worse.
The gap is that station-level metrics are readily available and routinely reported, while system-level flow efficiency is rarely measured. Without OSLFE, organizations have no way to detect when AI orchestration is locally optimizing at the expense of system flow. Every station dashboard can show green while the plant-level lead time is increasing and WIP inventory is growing. OSLFE makes the system-level reality visible by comparing total flow time to actual value-adding processing time.
Formula
OSLFE is calculated as the ratio of value-adding cycle times to total plant flow time:
OSLFE = Sum of Station_cycle_times / Plant_flow_time
Where:
- Station_cycle_times = sum of actual processing time at each work center for a unit or batch (minutes)
- Plant_flow_time = total elapsed time from release to completion for the same unit or batch (minutes)
A score of 1.0 means every minute in the plant is value-adding processing time, which is the theoretical ideal. Lower values indicate more time spent waiting, in transit, in queue, or in WIP buffers. A score of 0.1 means only 10% of the time a unit spends in the plant involves actual processing.
Key considerations for accurate measurement: Product mix complexity means different products have different ideal OSLFE values. Track by product family and compare to the pre-AI-orchestration baseline for the same family. Intentional WIP buffers (safety stock between stations) count as flow waste by OSLFE's definition but may be intentional for reliability. Track OSLFE both with and without planned buffer time to separate intentional from unintentional flow waste. For batch processes, use batch flow time rather than individual unit time. Minimum sample of 30 production units or batches per product family per measurement period for meaningful comparison.
Proposed formula, pending validation pilots.