Skip to content

BatchTrackingEvents¤

Track batch/recipe production from a batch-ID string signal.

Module: ts_shape.events.production.batch_tracking Guide: OEE Analytics

When to Use¤

Use for batch production environments (food, pharma, chemicals). Tracks batch boundaries from a batch-ID signal and computes duration and yield statistics. Essential for regulatory traceability and for optimizing batch scheduling to minimize transition losses.

Quick Example¤

from ts_shape.events.production.batch_tracking import BatchTrackingEvents
import pandas as pd
import numpy as np

df = pd.DataFrame({
    "timestamp": pd.date_range("2024-01-01 06:00", periods=360, freq="1min"),
    "batch_id": ["BATCH-001"]*90 + ["BATCH-002"]*120 + ["BATCH-003"]*150,
    "part_counter": np.arange(1, 361),
})

bt = BatchTrackingEvents(df, timestamp_column="timestamp")
batches = bt.detect_batches()
stats = bt.batch_duration_stats()
yields = bt.batch_yield(counter_uuid="part_counter")
matrix = bt.batch_transition_matrix()
print(batches)

Key Methods¤

Method Purpose Returns
detect_batches() Identify batch start and end times from value changes in the batch-ID signal DataFrame with batch_id, start, end, and duration
batch_duration_stats() Compute duration statistics (min, mean, max, std) grouped by batch type DataFrame with duration statistics per batch type
batch_yield(counter_uuid) Calculate production yield per batch using a counter signal DataFrame with batch_id and total parts produced
batch_transition_matrix() Build a matrix showing which batch follows which, useful for scheduling analysis DataFrame pivot table of from-batch vs to-batch counts

Tips & Notes¤

Watch for empty or null batch IDs

Gaps in the batch-ID signal (empty strings or nulls) often represent cleaning or changeover periods. Filter these out or treat them as explicit changeover events.

Related modules

See Also¤