Shift Reports & KPIs

Performance loss tracking, scrap costing, target comparison, setup time analysis (SMED), operator performance, rework tracking, and auto-generated shift handover reports.

---

Reporting Layer

These classes sit at the top of the production analytics stack — they consume data from tracking classes and produce actionable KPIs.

flowchart TB
    subgraph TRACKING["<b>Tracking Classes</b><br/><i>(see Production Monitoring guide)</i>"]
        DT["DowntimeTracking"]
        PT["PartProductionTracking"]
        QT["QualityTracking"]
        CT["CycleTimeTracking"]
    end

    subgraph KPI["<b>KPI & Loss Classes</b>"]
        PL["PerformanceLossTracking<br/><i>speed loss vs target</i>"]
        SC["ScrapTracking<br/><i>waste by reason & cost</i>"]
        TT["TargetTracking<br/><i>actual vs plan</i>"]
        ST["SetupTimeTracking<br/><i>SMED analysis</i>"]
        OP["OperatorPerformanceTracking<br/><i>operator comparison</i>"]
        RW["ReworkTracking<br/><i>rework rates & cost</i>"]
    end

    subgraph REPORT["<b>Reports</b>"]
        SH["ShiftHandoverReport<br/><i>auto-generated</i>"]
        PS["PeriodSummary<br/><i>weekly / monthly</i>"]
    end

    DT --> PL
    CT --> PL
    QT --> SC
    PT --> TT

    PL --> SH
    SC --> SH
    TT --> SH

    SH --> PS

    style TRACKING fill:#0f2a3d,stroke:#38bdf8,color:#e0f2fe
    style KPI fill:#1a3a4a,stroke:#f59e0b,color:#fef3c7
    style REPORT fill:#14532d,stroke:#22c55e,color:#dcfce7

---

Performance Loss Tracking

Identify when actual cycle times fall below the target — the "Performance" factor in OEE.

from ts_shape.events.production.performance_loss import PerformanceLossTracking

perf = PerformanceLossTracking(df, shift_definitions={
    "day": ("06:00", "14:00"),
    "afternoon": ("14:00", "22:00"),
    "night": ("22:00", "06:00"),
})

# Performance percentage per shift
by_shift = perf.performance_by_shift(
    cycle_uuid='cycle_trigger', target_cycle_time=30.0
)

# Identify hours where performance drops below 90%
slow = perf.slow_periods(
    cycle_uuid='cycle_trigger', target_cycle_time=30.0,
    threshold_pct=90.0, window='1h'
)

# Daily performance trend
trend = perf.performance_trend(
    cycle_uuid='cycle_trigger', target_cycle_time=30.0, window='1D'
)

---

Scrap Tracking

Track material waste by shift, reason code, and monetary cost.

from ts_shape.events.production.scrap_tracking import ScrapTracking

scrap = ScrapTracking(df, shift_definitions={
    "day": ("06:00", "14:00"),
    "afternoon": ("14:00", "22:00"),
    "night": ("22:00", "06:00"),
})

# Scrap quantity per shift
by_shift = scrap.scrap_by_shift(scrap_uuid='scrap_counter')

# Pareto of scrap reasons
by_reason = scrap.scrap_by_reason(
    scrap_uuid='scrap_counter', reason_uuid='scrap_reason'
)

# Convert to monetary cost
cost = scrap.scrap_cost(
    scrap_uuid='scrap_counter',
    part_id_uuid='part_number',
    material_costs={'PART_A': 12.50, 'PART_B': 8.75}
)

# Scrap trend over time
trend = scrap.scrap_trend(scrap_uuid='scrap_counter', window='1D')

---

Target Tracking

Compare actual production metrics against shift or daily targets.

from ts_shape.events.production.target_tracking import TargetTracking

targets = TargetTracking(df, shift_definitions={
    "day": ("06:00", "14:00"),
    "afternoon": ("14:00", "22:00"),
    "night": ("22:00", "06:00"),
})

# Actual vs target per shift
comparison = targets.compare_to_target(
    metric_uuid='part_counter',
    targets={'day': 500, 'afternoon': 500, 'night': 400}
)

# Daily target achievement summary
summary = targets.target_achievement_summary(
    metric_uuid='part_counter', daily_target=1400
)

# Target hit rate (% of days meeting target)
hit_rate = targets.target_hit_rate(
    metric_uuid='part_counter', daily_target=1400
)

---

Setup Time Tracking (SMED)

Analyze changeover/setup durations for SMED (Single-Minute Exchange of Die) improvement.

from ts_shape.events.production.setup_time_tracking import SetupTimeTracking

setup = SetupTimeTracking(df, shift_definitions={
    "day": ("06:00", "14:00"),
    "afternoon": ("14:00", "22:00"),
    "night": ("22:00", "06:00"),
})

# List every setup event with duration
durations = setup.setup_durations(state_uuid='machine_state', setup_value='Setup')

# Setup time by product transition (from → to)
by_product = setup.setup_by_product(
    state_uuid='machine_state', part_id_uuid='part_number'
)

# Overall setup statistics
stats = setup.setup_statistics(state_uuid='machine_state')

# Track setup time improvement over time
trend = setup.setup_trend(state_uuid='machine_state', window='1W')

---

Operator Performance Tracking

Compare production output and quality across operators or teams.

from ts_shape.events.production.operator_performance import OperatorPerformanceTracking

ops = OperatorPerformanceTracking(df, shift_definitions={
    "day": ("06:00", "14:00"),
    "afternoon": ("14:00", "22:00"),
    "night": ("22:00", "06:00"),
})

# Parts produced per operator
production = ops.production_by_operator(
    operator_uuid='operator_id', counter_uuid='part_counter'
)

# Operator efficiency vs target
efficiency = ops.operator_efficiency(
    operator_uuid='operator_id', counter_uuid='part_counter',
    target_per_shift=500
)

# Quality (First Pass Yield) per operator
quality = ops.quality_by_operator(
    operator_uuid='operator_id',
    ok_uuid='good_parts',
    nok_uuid='bad_parts'
)

# Ranked operator comparison
ranking = ops.operator_comparison(
    operator_uuid='operator_id', counter_uuid='part_counter'
)

---

Rework Tracking

Track parts that require rework — re-processing through a station.

from ts_shape.events.production.rework_tracking import ReworkTracking

rework = ReworkTracking(df, shift_definitions={
    "day": ("06:00", "14:00"),
    "afternoon": ("14:00", "22:00"),
    "night": ("22:00", "06:00"),
})

# Rework count per shift
by_shift = rework.rework_by_shift(rework_uuid='rework_counter')

# Rework by reason code
by_reason = rework.rework_by_reason(
    rework_uuid='rework_counter', reason_uuid='rework_reason'
)

# Rework rate as % of total production
rate = rework.rework_rate(
    rework_uuid='rework_counter',
    total_production_uuid='total_counter'
)

# Convert to monetary cost
cost = rework.rework_cost(
    rework_uuid='rework_counter',
    part_id_uuid='part_number',
    rework_costs={'PART_A': 5.00, 'PART_B': 3.50}
)

# Rework trend
trend = rework.rework_trend(rework_uuid='rework_counter', window='1D')

---

Shift Handover Report

Auto-generate shift handover reports from raw signals — production, quality, downtime, and issues.

from ts_shape.events.production.shift_handover import ShiftHandoverReport

handover = ShiftHandoverReport(df, shift_definitions={
    "day": ("06:00", "14:00"),
    "afternoon": ("14:00", "22:00"),
    "night": ("22:00", "06:00"),
})

# Generate full report from raw signals
report = handover.generate_report(
    counter_uuid='part_counter',
    ok_counter_uuid='good_parts',
    nok_counter_uuid='bad_parts',
    state_uuid='machine_state',
    targets={'day': 500, 'afternoon': 500, 'night': 400},
    quality_target_pct=98.0,
    availability_target_pct=90.0,
    report_date='2024-01-15'
)

# Highlight issues requiring attention
issues = handover.highlight_issues(
    counter_uuid='part_counter',
    ok_counter_uuid='good_parts',
    nok_counter_uuid='bad_parts',
    state_uuid='machine_state',
    report_date='2024-01-15'
)

Build from pre-computed data

report = ShiftHandoverReport.from_shift_data(
    production_df=production_by_shift,
    quality_df=quality_by_shift,
    downtime_df=downtime_by_shift,
    targets={'day': 500, 'afternoon': 500, 'night': 400},
    report_date='2024-01-15'
)

---

Period Summary

Roll up daily data to weekly or monthly aggregations.

from ts_shape.events.production.period_summary import PeriodSummary

summary = PeriodSummary(df)

# Weekly summary
weekly = summary.weekly_summary(
    counter_uuid='part_counter',
    ok_counter_uuid='good_parts',
    nok_counter_uuid='bad_parts'
)

# Monthly summary
monthly = summary.monthly_summary(
    counter_uuid='part_counter',
    ok_counter_uuid='good_parts',
    nok_counter_uuid='bad_parts'
)

# Compare two periods
comparison = summary.compare_periods(
    counter_uuid='part_counter',
    period1=('2024-01-01', '2024-01-31'),
    period2=('2024-02-01', '2024-02-29')
)

Roll up from pre-computed daily data

monthly = PeriodSummary.from_daily_data(
    daily_df=daily_production, freq='M', date_column='date'
)

---

Module Deep Dives

Setup & Operators: Setup Time (SMED) | Operator Performance | Rework Tracking

Performance & Reporting: Performance Loss | Scrap Tracking | Target Tracking | Shift Handover | Period Summary

---

Next Steps

• Production Monitoring — The tracking classes that feed into these reports
• OEE & Plant Analytics — OEE calculation from the same signals
• Product Traceability — Track individual parts through the process
• API Reference — Full reporting API documentation