Bottleneck Identification & Protection Workflow

Purpose

Find the constraint (bottleneck) that limits system throughput, measure its real-world effective capacity, confirm it is the true constraint, and protect it so the whole line can produce more reliably.

Quick Recipe (high level)

1. Map the flow and identify candidate bottlenecks.
2. Measure effective capacity for each candidate (real output over available time).
3. Run simple constraint tests to confirm the system constraint.
4. Set protection rules around the constraint (WIP limits upstream; explicit dispatch/priority rules).
5. Monitor key indicators and run short experiments to verify throughput gain.

Step 1 — Map the flow and find candidates

Draw the value stream at the level you operate (cell, line, or plant). For each process step capture:

• Cycle time (typical)
• Available time per shift/day
• Observed uptime/downtime
• Average queue length and WIP before the step

Candidate bottlenecks are steps with long cycle times, frequent downtime, large queues, or that visibly limit downstream flow.

Step 2 — Measure effective capacity

Effective capacity = (Available time) × (Observed uptime fraction) ÷ (Average cycle time per unit).

Practical measurement approach:

1. Choose a representative short window (2–4 shifts).
2. Record units completed at the station and the station’s available time during that window.
3. Estimate uptime (available - downtime) or use existing downtime logs.
4. Compute units/hour or units/shift as the effective capacity.

Rank steps by measured effective capacity versus required demand to see which step constrains output.

Step 3 — Constraint tests (simple, low-risk)

Two lightweight tests to confirm the true system constraint:

• Starve test: Temporarily hold upstream WIP to the suspected station (reduce WIP entering it). If overall throughput drops, the station is likely the constraint.
• Load test: Move extra work onto the suspected station (add work-in from upstream). If throughput increases, it supports the station being the constraint.

Run tests for a full production cycle (or at least a shift) and measure throughput, downstream blocking, and queue lengths. Use conservative changes and pre-agreed safety/quality guardrails.

Step 4 — Set protection rules

Protect the constraint with clear, enforceable rules so it works without interruption:

• Upstream WIP limits: Set a fixed maximum WIP in front of the constrained station to prevent starvation and excessive queues. Use Little's Law to estimate: WIP ≈ throughput × lead time. Start conservative and adjust by experiment.
• Priority/dispatch rules: Define which items get priority at the constrained station (e.g., next due date, highest value, customer priority). Keep rules simple and visible at the line.
• Buffer management: Provide a small inbound buffer with clear replenishment rules. Use visual signals (kanban) or digital flags.
• Maintenance & uptime focus: Schedule constraint-focused preventive maintenance, quick repair packs, and trained operators who can troubleshoot fast.

Step 5 — Continuous monitoring recipe

Track a small set of KPIs to know whether protections are working:

• Throughput (units per shift or hour) — primary metric
• Constraint utilization and uptime
• Average queue length / WIP upstream of constraint
• Blocking incidents (downstream starve/block events)
• Cycle time variance at the constraint

Set simple control limits (e.g., daily throughput target ± x%) and automatic alerts or huddles when breached.

Experiment & verification tracker (use for short improvement cycles)

For each experiment use the following fields and record results for at least one shift to gather evidence:

1. Experiment name
2. Goal (e.g., increase throughput by 8% without raising total WIP)
3. Change (what you will change: new WIP limit, priority rule, minor staffing shift)
4. Duration (start/end shifts or dates)
5. Success criteria (quantitative: throughput increase, reduced blocking; qualitative: no quality regressions)
6. Measurements collected (throughput, constraint uptime, upstream WIP, defects)
7. Outcome & next step (adopt, adapt, roll-back, or iterate)

Common pitfalls and how to avoid them

• Optimizing non-constraint steps: Avoid local efficiencies that create more WIP and greater variability. Prioritize work on the constraint.
• Too-large buffers: Excessive upstream WIP hides problems and increases lead time. Use the smallest buffer that keeps the constraint fed.
• Ignoring variation: Capacity estimates must account for downtime and variability — use observed uptime and cycle time distributions, not ideal cycle times.
• No defined ownership: Assign a clear owner for the constraint protection rules and daily checks.

Next practical steps

1. Run the mapping and measure effective capacity for candidate stations this week.
2. Schedule one short, low-risk starve or load test to confirm the constraint.
3. Define simple protection rules and run a 3–5 shift experiment with the tracker fields above.
4. Hold a quick experiment review huddle, adopt what works, and document the protection rules in standard work.