Lean Manufacturing and Operational Metrics

Overall Equipment Effectiveness (OEE)

OEE is the single most widely used metric for measuring manufacturing productivity. It is defined as the product of three factors:

OEE = Availability × Performance × Quality

• Availability = (Planned production time − Unplanned downtime) ÷ Planned production time. Captures unplanned stops: breakdowns, material shortages, operator absence.
• Performance = (Actual output × Ideal cycle time) ÷ Actual run time. Captures speed losses: running slower than the design rate, minor stops, micro-stoppages.
• Quality = Good parts ÷ Total parts started. Captures defects: scrap, rework, and first-pass yield losses.

Worked example: A press runs 8-hour shifts. Planned production time is 420 minutes (8 hours minus 60 minutes of scheduled breaks). Unplanned downtime (breakdowns + material shortages) totalled 60 minutes. The press has an ideal cycle time of 0.5 minutes per part. Total parts produced: 600. Good parts: 570 (30 scrap).

• Availability = (420 − 60) ÷ 420 = 360 ÷ 420 = 85.7%
• Performance = (600 × 0.5) ÷ 360 = 300 ÷ 360 = 83.3%
• Quality = 570 ÷ 600 = 95.0%
• OEE = 0.857 × 0.833 × 0.950 = 67.8%

The widely cited "world-class OEE" benchmark is 85%, per the OEE Foundation and the original OEE research by Nakajima (1988). A typical starting OEE for manufacturers beginning to measure it is 35–65%. The 85% benchmark is for mature, high-volume repetitive production — a 75% OEE in a high-variety job shop may be excellent in context.

Use the OEE calculator to compute your own figure with a full loss breakdown.

Takt time, cycle time and lead time: the differences

These three terms are routinely confused, including by people who should know better:

Worked example: You work one shift of 450 minutes per day (480 min shift minus 30 min breaks). Customer demand is 90 units per day.

• Takt time = 450 ÷ 90 = 5 minutes per unit
• Any workstation with a cycle time > 5 minutes is a bottleneck — it cannot keep pace with demand.
• A workstation with a cycle time of 3 minutes has spare capacity that could be used to support the bottleneck.

Kanban for small factories

Kanban is a pull-based scheduling method that controls work-in-process (WIP) by limiting the number of items that can be in production at any one time. The name comes from the Japanese word for card or signal — originally, physical cards were attached to containers of parts. When a downstream process consumed a container, the card signalled the upstream process to replenish it.

The number of kanban cards (or signals) in the system controls WIP and smooths flow:

Number of kanbans = (D × LT × (1 + SS)) ÷ Q

Where D = average demand per period, LT = replenishment lead time (in same units as D), SS = safety stock factor (typically 0.1–0.2), Q = container/batch size.

Kanban works well for: repetitive production with stable demand, short production runs between changeovers, and where visual management is valuable. It works poorly for: high-mix low-volume job shops, items with highly variable demand, or processes with very long lead times.

5S workplace organisation

5S is a structured approach to workplace organisation derived from five Japanese words: Seiri (Sort), Seiton (Set in order), Seiso (Shine), Seiketsu (Standardise), Shitsuke (Sustain). In English:

1. Sort — remove everything from the workplace that is not needed for current work. Red-tag items of uncertain usefulness; dispose of confirmed waste.
2. Set in order — assign a specific, marked location to every item that remains. "A place for everything, and everything in its place."
3. Shine — clean the workplace to its operational standard. Cleaning is also inspection — dirt and wear become visible.
4. Standardise — document the standard condition with checklists, shadow boards and visual controls. Define who does what and when.
5. Sustain — audit to the standard regularly. The hardest step: without management discipline, 5S deteriorates within weeks.

5S is often the starting point for lean because it creates the visual baseline from which improvements can be seen and abnormalities detected. A 5S audit score (percentage of audit points at standard) is a leading indicator of general process discipline.

SMED: reducing changeover time

Single-Minute Exchange of Die (SMED) is Shigeo Shingo's methodology for reducing the time to change over a machine from one product to another. The "single-minute" target means changeover in under 10 minutes (a single digit of minutes). The core principle is converting internal setup (done while the machine is stopped) to external setup (done while the machine is still running).

The SMED process in four steps:

1. Observe and document — video the entire changeover. Time every element. Most teams discover they have never actually timed a changeover accurately before.
2. Separate internal and external elements — which steps require the machine to be stopped? Which can be done in advance or after restart? Typically 30–50% of changeover elements can be externalised with no capital investment.
3. Convert internal to external — pre-stage tooling, dies, materials and settings before the machine stops. This alone typically cuts changeover time by 30–50%.
4. Streamline remaining internal elements — standardise clamping mechanisms, use quick-release fasteners, eliminate adjustments by using pre-set tooling, parallelize operations where two people can work simultaneously.

SMED is most valuable for: stamping, injection moulding, extrusion, CNC machining, food processing — any process where a long changeover drives large batch sizes and high WIP inventory.

Value Stream Mapping (VSM)

A value stream map is a diagram that shows the flow of material and information required to bring a product from raw material to customer. It shows: process steps, inventory between steps, cycle times, changeover times, uptime, number of operators, information flows, and push vs pull signals.

The current-state map reveals waste; the future-state map designs the improved flow. VSM is a planning tool, not a monitoring tool — it is done in a team workshop (typically 1–2 days) and updated periodically as improvement projects are completed.

VSM is most useful for: identifying the dominant sources of lead time (usually inventory between processes, not process time), prioritising where to focus improvement projects, and communicating the current and future state visually to teams and management.