Workforce Analytics

Overview

DNSPs employ large workforces of skilled electrical tradespeople, engineers, and contractors. Workforce costs typically represent 25–35% of total opex. Efficient workforce management — having the right skills, in the right place, at the right time — is both a cost efficiency imperative and an AER benchmarking requirement.

Energy Copilot’s Workforce Analytics module provides a data-driven view of workforce performance, capacity, and forward demand.

ML Demand Forecasting

The workforce demand forecasting model (Prophet + XGBoost ensemble) predicts the number of field crew hours required across skill categories over an 18-month horizon.

Model Architecture

Prophet (trend + seasonality) ──┐
                                 ├──→ Ensemble → Forecast
XGBoost (exogenous features) ───┘

Prophet components:

Long-term trend (workforce growth with network expansion)
Annual seasonality (higher demand in summer for storm response, lower in winter)
Weekly seasonality (Monday peak for planned work, reduced weekends)

XGBoost features:

Capex program spend (drives contractor demand for large projects)
Planned vegetation clearance program (seasonal)
Scheduled asset replacements
Storm event indicator (drives unplanned demand)
Public holiday flag

Model Performance

Metric	Value
MAE (monthly)	124 crew-hours
MAPE	4.2%
Forecast horizon	18 months
Retraining frequency	Monthly

See Workforce Demand Forecasting ML for technical details.

Dashboard Pages

Workforce Demand Forecast (`/dnsp/workforce/forecast`)

18-month forecast chart by skill category (HV/LV linesperson, cable jointer, engineer, project manager)
Comparison: forecast vs plan (from annual workforce plan)
Gap analysis: forecast demand minus current capacity = gap to fill via contractors
Confidence intervals (P10/P50/P90)

Screenshot: Workforce demand forecast chart showing 18-month horizon with actual (solid), P50 forecast (dashed), and P10/P90 confidence band (shaded).

Contractor Scorecard (`/dnsp/workforce/contractors`)

Performance tracking for external contractors:

Metric	Calculation	Target
Job completion rate	Jobs completed on time / Total jobs	>95%
Defect rate	Defective work items / Total work items	<2%
Safety record	LTI frequency rate	<1.5
Cost efficiency	Actual cost / Quoted cost	<1.05
Response time	Hours from callout to arrival (emergency)	<4h

Field Crew Productivity (`/dnsp/workforce/productivity`)

Productivity metrics by work type and crew:

-- Crew productivity by work type
SELECT
    crew_id,
    work_type,
    region,
    completed_jobs,
    average_job_duration_h,
    travel_ratio_pct,   -- travel time / total time
    utilisation_pct,    -- productive time / available time
    productivity_index  -- actual output / expected output
FROM energy_copilot.gold.dnsp_workforce_metrics
WHERE period_month = DATE_TRUNC('month', CURRENT_DATE())
ORDER BY productivity_index ASC  -- show lowest performers first
LIMIT 20;

Skills Gap Analysis (`/dnsp/workforce/skills-gap`)

Skills gap analysis compares current skills inventory against forecast demand:

Skill	Current Headcount	Forecast Peak Demand (12m)	Gap	Gap Type
HV Linesperson	85	92	-7	Shortage
Cable Jointer	32	28	+4	Surplus
Protection Engineer	12	15	-3	Shortage
Project Manager	18	20	-2	Shortage

Skills gaps drive recruitment targets and contractor pre-qualification requirements.

Bushfire Seasonality

Bushfire season creates significant demand spikes for vegetation management crews:

Nov-Feb: +35% to +80% above base demand (vegetation clearance emergency response)
Mar-Apr: +15% above base (post-season remediation)
May-Oct: base demand

The forecast model explicitly incorporates bushfire season effects, and the workforce plan configures standby contractor arrangements for the peak period.

Opex Benchmarking

Workforce opex is benchmarked against AER efficiency benchmarks and peer DNSPs:

# Workforce opex benchmark comparison
GET /api/dnsp/workforce/opex-benchmark?dnsp=ausgrid&year=2024

# Response includes:
# - total_workforce_opex_aud
# - per_customer_aud (benchmarked vs peer group)
# - per_km_network_aud (benchmarked vs peer group)
# - efficiency_score (0-100, 100 = most efficient in peer group)
# - aer_reference_expenditure_aud (AER's approved opex allowance)

API Endpoints

# Workforce demand forecast
GET /api/dnsp/workforce/forecast?dnsp=ergon&horizon=18m

# Contractor scorecard
GET /api/dnsp/workforce/contractors?dnsp=sa_power_networks&period=2025-Q1

# Skills gap analysis
GET /api/dnsp/workforce/skills-gap?dnsp=essential_energy

# Crew productivity
GET /api/dnsp/workforce/productivity?dnsp=energex&work_type=vegetation

# Forecast vs plan comparison
GET /api/dnsp/workforce/plan-comparison?dnsp=ausgrid&year=2025