Skip to content

InventoryMonitoringEvents¤

Monitor inventory levels from a timeseries signal and detect supply chain events.

Module: ts_shape.events.supplychain.inventory_monitoring Guide: Production Guide

When to Use¤

Use for real-time inventory monitoring in manufacturing. Tracks stock levels, calculates consumption rates, and predicts when a stockout will occur based on current consumption trends. Works with any signal representing a quantity that depletes over time and is periodically replenished.

Quick Example¤

from ts_shape.events.supplychain.inventory_monitoring import InventoryMonitoringEvents

monitor = InventoryMonitoringEvents(
    df=inventory_df,
    timestamp_col="timestamp",
    value_col="stock_level"
)

# Flag intervals where stock drops below safety threshold
low_stock = monitor.detect_low_stock(threshold=100, min_duration="2H")
print(f"Low stock events: {len(low_stock)}")

# Calculate rolling consumption rate
rate = monitor.consumption_rate(window="1D")

# Detect when stock falls below the reorder point
breaches = monitor.reorder_point_breach(reorder_point=250)

# Predict time until stockout at current consumption rate
prediction = monitor.stockout_prediction(window="7D")
if prediction is not None:
    print(f"Predicted stockout in: {prediction.days} days")

Key Methods¤

Method Purpose Returns
detect_low_stock() Flag low inventory intervals DataFrame of periods below stock threshold
consumption_rate() Rolling consumption rate DataFrame with consumption rate per window
reorder_point_breach() Detect when stock falls below reorder point DataFrame of breach events with timestamps
stockout_prediction() Time estimate until stockout Timedelta estimate or None if stock is stable

Tips & Notes¤

Account for replenishment events

Stock level jumps (replenishments) can skew consumption_rate(). Filter out positive deltas or use a median-based rate to get robust consumption estimates that ignore restocking spikes.

Related modules

See Also¤