Skip to content

FlowConstraintEvents¤

Detect blocked and starved conditions between upstream/downstream stations.

Module: ts_shape.events.production.flow_constraints Guide: Production Monitoring

When to Use¤

Use for multi-station line analysis. Identifies where material flow is constrained — blocked (downstream can't keep up) or starved (upstream can't supply). Understanding flow constraints is critical for line balancing and identifying the true bottleneck versus symptomatic idle time.

Quick Example¤

from ts_shape.events.production.flow_constraints import FlowConstraintEvents
import pandas as pd

df = pd.DataFrame({
    "timestamp": pd.date_range("2024-01-01 06:00", periods=240, freq="30s"),
    "station_1_running": [True]*50 + [True]*40 + [False]*20 + [True]*130,
    "station_2_running": [True]*50 + [False]*40 + [False]*20 + [True]*130,
})

roles = {"upstream": "station_1_running", "downstream": "station_2_running"}
fc = FlowConstraintEvents(df, timestamp_column="timestamp")
blocked = fc.blocked_events(roles=roles, tolerance="200ms")
starved = fc.starved_events(roles=roles, tolerance="200ms")
print(blocked.head())

Key Methods¤

Method Purpose Returns
blocked_events(roles, tolerance='200ms') Detect intervals where upstream is running but downstream is stopped (blocked condition) DataFrame of blocked intervals with start, end, and duration
starved_events(roles, tolerance='200ms') Detect intervals where downstream is idle due to no upstream supply (starved condition) DataFrame of starved intervals with start, end, and duration
flow_constraint_analytics(roles) Comprehensive analytics combining blocked, starved, and free-flow statistics Dict with summary metrics and interval DataFrames

Tips & Notes¤

Use tolerance to handle signal jitter

Set tolerance to account for PLC scan-time differences between stations — 200ms to 1s is typical. Without tolerance, you may get false blocked/starved detections from timing mismatches.

Related modules

See Also¤