Key Performance Indicators¶

Core metrics for measuring warehouse performance.

What are KPIs?¶

Key Performance Indicators (KPIs) are the critical metrics that define success. They summarize complex system behavior into actionable numbers.

Throughput KPIs¶

Tasks Per Hour¶

Primary throughput measure:

Throughput = Tasks Completed / Hours

Target: Depends on warehouse requirements

Example:

1,250 tasks in 1 hour
Throughput = 1,250 tasks/hour

Orders Per Hour¶

Order-level throughput:

Order Throughput = Orders Completed / Hours

Note: One order may have multiple tasks.

Peak Throughput¶

Maximum sustainable rate:

Peak Throughput = Max tasks/hour over sustained period

Time KPIs¶

Average Task Time¶

Mean time from creation to completion:

Avg Task Time = Σ(completion_time - creation_time) / task_count

Components:

Task Time = Queue Wait + Travel + Traffic Wait + Station Queue + Service

Order Cycle Time¶

Time to fulfill complete order:

Cycle Time = Order Completion - Order Creation

P95 Task Time¶

95^th percentile task time:

95% of tasks complete within this time

Captures worst-case performance better than average.

Utilization KPIs¶

Robot Utilization¶

How busy robots are:

Robot Utilization = (Working + Traveling) / Total Time

Breakdown:

100% = Working + Traveling + Waiting + Idle + Charging + Maintenance

Target: 70-85% (too high causes congestion)

Station Utilization¶

How busy stations are:

Station Utilization = Service Time / Total Time

Target: Varies by station type

Fleet Efficiency¶

Productive work vs. overhead:

Fleet Efficiency = Work Time / (Work Time + Travel Time + Wait Time)

Quality KPIs¶

On-Time Delivery Rate¶

Tasks meeting due time:

On-Time Rate = On-Time Tasks / Total Tasks × 100%

Target: 95%+ typically

Late Task Count¶

Absolute count of missed due times:

Late Tasks = Tasks where completion > due_time

Average Lateness¶

For late tasks only:

Avg Lateness = Σ(completion_time - due_time) / late_task_count

Congestion KPIs¶

Average Wait Time¶

Time robots spend waiting:

Avg Wait = Total Wait Time / Task Count

Congestion Rate¶

Percentage of time edges/nodes are at capacity:

Congestion Rate = Time at Capacity / Total Time

Deadlock Count¶

Number of deadlocks detected:

Deadlocks = Count of deadlock events

Target: 0 ideally

Reliability KPIs¶

Robot Availability¶

Fleet uptime:

Availability = (Operating + Idle) / Total Time

Mean Time Between Failures¶

Average operating time before failure:

MTBF = Total Operating Time / Failure Count

Mean Time To Repair¶

Average repair duration:

MTTR = Total Repair Time / Failure Count

KPI Dashboard¶

Summary View¶

┌─────────────────────────────────────────────────┐
│                  KPI DASHBOARD                   │
├─────────────────────────────────────────────────┤
│ Throughput                                       │
│   Tasks/Hour:     1,250    (Target: 1,000) ✓    │
│   Orders/Hour:      312    (Target: 250)  ✓    │
│                                                  │
│ Time                                             │
│   Avg Task Time:   42.3s   (Target: 60s)  ✓    │
│   P95 Task Time:   85.0s   (Target: 120s) ✓    │
│                                                  │
│ Utilization                                      │
│   Robot:           78%     (Target: 70-85%)✓    │
│   Station:         82%     (Target: 80%)  ✓    │
│                                                  │
│ Quality                                          │
│   On-Time Rate:    96.5%   (Target: 95%)  ✓    │
│   Late Tasks:      44      (Target: <100) ✓    │
│                                                  │
│ Congestion                                       │
│   Avg Wait:        3.2s    (Target: <5s)  ✓    │
│   Deadlocks:       0       (Target: 0)    ✓    │
└─────────────────────────────────────────────────┘

Setting KPI Targets¶

Baseline First¶

Run baseline simulation:

waremax run baseline.yaml -o baseline/

Extract current KPIs as starting point.

Industry Benchmarks¶

KPI	Good	Excellent
Robot utilization	70%	80%
On-time rate	95%	99%
P95 task time	2× average	1.5× average

Business Requirements¶

Work backward from business needs:

Need: 1,000 orders/day
Hours: 16 operating hours
Required: 62.5 orders/hour

Monitoring KPIs¶

Real-Time Tracking¶

metrics:
  timeseries:
    enabled: true
    interval_s: 60
    include:
      - throughput
      - utilization
      - queue_length

Alerting Thresholds¶

metrics:
  alerts:
    throughput_below: 800
    utilization_above: 90
    wait_time_above: 10.0

KPI Trade-offs¶

Throughput vs. Utilization¶

High utilization → More congestion → Lower effective throughput

Find balance point.

Speed vs. Quality¶

Faster service → Less thorough → Potential errors

Cost vs. Performance¶

More robots → Higher throughput → Higher cost

Configuration¶

metrics:
  kpis:
    enabled: true
    include:
      - throughput
      - task_time
      - utilization
      - on_time_rate
      - wait_time

  targets:
    throughput_per_hour: 1000
    avg_task_time_s: 60
    on_time_rate_pct: 95