Bottleneck Detector is the master sensor that answers:
Which OS layer is limiting the whole system right now?
Most systems fail because they keep repairing the wrong layer.
They push harder on the visible surface symptom, while the real bottleneck sits deeper.
Bottleneck Detector prevents wasted effort and stops drift.
What This Sensor Measures
Bottleneck Detector measures:
- constraint dominance
- which layer is currently binding and preventing growth
- downstream symptoms
- which failures appear above it (secondary effects)
- repair leverage
- which fix would produce the largest recurrence drop
- cross-sensor disagreement
- which sensor panel shows the sharpest weakness versus the rest
- time-to-stability
- which layer takes longest to stabilise (usually the bottleneck)
A bottleneck is not “the biggest problem”.
A bottleneck is the limiting factor.
How to Read This Sensor
Bottleneck is clear when:
- one sensor cluster is obviously weaker than others
- repairs elsewhere don’t change outcomes
- recurrence remains high
- slope stays flat
- the same failure returns despite effort
Bottleneck is hidden when:
- multiple problems appear simultaneously
- narratives dominate (drift high)
- coordination cost blocks repair propagation
- measurement is weak
When bottleneck is hidden, the first repair is always:
turn on sensors + reduce drift.
Minimum Viable Test (So Bottleneck Can Be Detected)
A system passes the minimum bottleneck test when it can:
- list top 3 recurring failures
- map each failure to a layer
- identify which layer, if fixed, would drop recurrence fastest
- confirm via a short repair experiment
If you can’t map failures to layers, you can’t detect bottleneck.
Bottleneck Categories (Which Layer Can Be the Limiter)
A) Mind OS bottleneck
- attention unstable
- judgement noisy
- regulation low
- learning and repair collapse under stress
B) Vocabulary OS bottleneck
- meaning friction
- retrieval weak
- wrong usage recurrence
- comprehension and writing limited
C) Education OS bottleneck
- comprehension slope flat
- low transfer
- slow learning velocity
- repeated errors
D) Governance OS bottleneck
- execution friction
- accountability weak
- coordination cost rising
E) Production OS bottleneck
- no surplus
- buffers thin
- maintenance backlog grows
F) Constraint OS bottleneck
- ceilings binding
- dependency risk high
- buffers exhausted
G) Drift bottleneck
- truth failure
- metrics gamed
- feedback suppressed
H) Repair bottleneck
- recurrence not dropping
- patches not holding
- repair does not propagate
Usually one dominates.
Bottleneck Levels (Gauge Alignment)
Level 1 — Clear Bottleneck
- one limiting layer obvious
- repair leverage is high
Level 2 — Probable Bottleneck
- two candidates
- short test needed
Level 3 — Mixed Bottlenecks
- multiple layers weak
- repair must prioritise foundations first
Level 4 — Hidden Bottleneck (Drift/Friction)
- measurement unclear
- truth flow distorted
- execution blocked
Level 5 — Bottleneck Cascade
- constraints + CDI + drift + coordination rising together
- system near brittleness zone
The One Mistake This Sensor Prevents
The most common failure is local optimisation:
- improving surface outcomes
- while recurrence stays high
- because the real bottleneck never changed
Bottleneck Detector forces:
fix the limiter, not the symptom.
Repair Actions (When Bottleneck Is Identified)
- stop working on non-bottleneck layers
- target the bottleneck
- define recurrence drop criteria
- retest after delay
- propagate the fix
That’s how systems scale.
Links (Planet OS Instrument Layer)
- Planet OS Sensors Directory: https://edukatesg.com/planet-os-sensors/
- Recurrence Index: https://edukatesg.com/planet-sensor-recurrence/
- Planet OS Dashboard: https://edukatesg.com/planet-os-dashboard/
- Planet OS Failure Modes: https://edukatesg.com/planet-os-failure-modes/
Next Page to Publish
Planet Sensor: Repair Capacity Growth (Canonical)
https://edukatesg.com/planet-sensor-repair-capacity-canonical/