OPERA PVBLICAPublic works
Infrastructure & Utilities
Networks, water systems and transport assets are judged over decades. We train their leaders to apply AI to condition, capital planning and resilience, with decisions regulators and the public can follow.
The landscape
Infrastructure is judged over decades and scrutinized every day. Utilities and public-works owners carry aging assets, larger capital programs, tighter regulation and rising expectations for resilience against storms, heat, wildfire and cyberattack. AI can show the condition of networks of pipe, wire, track and road, and direct capital to where risk is highest. It also brings new obligations: decisions regulators can follow, secure operational technology and fair treatment of the public being served. Arclight trains leaders of networks, water systems and transport assets to apply AI with the accountability their mandate requires.
- I
Aging assets, larger programs
Many networks include assets near or past their design life, while capital programs grow to meet demand, resilience and mandated replacement. Owners must decide what to repair, replace or retire with incomplete condition data.
- II
Resilience and liability
Wildfire, flood, extreme heat and storms now shape both capital plans and legal exposure. Regulators, insurers and courts expect evidence that risk models are sound and that mitigation spending is targeted where it reduces harm.
- III
Affordability, workforce and security
Rate increases and public budgets face scrutiny from regulators, councils and customers. Experienced operators are retiring, and operational technology faces persistent cyber threats, including from state-sponsored actors.
Where AI earns its place.
Six applications with proven economics in this sector, taught with the techniques behind them and the ways they fail.
- 01
Asset condition and risk scoring
Combine inspection records, sensor data, failure history and environmental conditions to score the likelihood and consequence of failure for each asset. Capital goes to the pipes, poles, bridges and substations where risk is highest.
- Gradient-boosted models
- Survival analysis
- Bayesian inference
- 02
Automated inspection
Analyze drone, LiDAR, vehicle and satellite imagery to detect corrosion, cracking, ground movement, encroachment and vegetation risk. Inspectors focus on confirmed defects, and the record stays consistent across contractors and years.
- Computer vision
- LiDAR point-cloud analysis
- SAR interferometry
- 03
Capital plan optimization
Select and sequence projects across a multi-year program under budget, outage, crew and regulatory constraints. Leaders see the risk reduced per dollar spent and can defend the plan in a rate case or budget hearing.
- Mixed-integer optimization
- Monte Carlo simulation
- 04
Wildfire, storm and flood resilience
Model weather-driven risk to each circuit, segment or structure, and plan hardening, vegetation management and protective shutoffs accordingly. Decisions are documented for the regulators, insurers and courts who will review them.
- Geospatial ML
- Probabilistic weather forecasting
- Gradient-boosted models
- 05
Network operations and leak detection
Detect leaks, bursts and abnormal pressure or flow in water and gas networks, and forecast demand to optimize pumping and storage. Operators respond sooner and reduce non-revenue water and energy cost.
- Anomaly detection
- Digital twins
- Time-series forecasting
- 06
Permitting, compliance and customer service
Draft permit applications and regulatory filings, answer customer questions and triage service requests with citation-grounded language models. Staff review and approve every output, and the record remains auditable.
- Retrieval-augmented generation
- LLM agents
- Document intelligence
The Infrastructure curriculum.
Taught within the Command Intensive, as an on-site immersion, or privately in one-to-one advisory.
- I
AI for long-lived assets
Where AI improves condition knowledge, planning and operations for assets with decades-long lives, and where better data collection matters more than better models. Participants map use cases to safety, service and cost.
- II
Risk-based capital planning
Building likelihood-of-failure and consequence models that regulators and auditors can follow. Executives learn to question model outputs, set confidence thresholds and connect asset risk to the capital plan.
- III
Resilience and operational technology
Applying AI to wildfire, storm, flood and heat risk, and deploying models near SCADA and control systems safely. Covers cyber exposure, human oversight and EU AI Act obligations for critical infrastructure where they apply.
- IV
Public accountability and governance
Explainable decisions for rate cases and public budgets, procurement of AI systems, workforce transition and the governance that keeps decisions defensible. The module ends with a board-ready roadmap.
Who attends
- Utility CEOs and general managers
- CFOs and heads of regulatory affairs
- COOs and heads of network operations
- Chief engineers and asset management directors
- Heads of capital delivery
- Public works and transport agency executives
Outcomes
- A risk-based AI portfolio tied to safety, reliability and affordability
- Decision methods that regulators, auditors and the public can follow
- Governance for AI in operational technology and critical systems
- A multi-year roadmap aligned to the capital program and rate cycle
How infrastructure owners engage us.
- IThe Command IntensiveIMPERIVMThree days to set the AI agenda for the enterprise.3 daysUp to 12 executives
- IIOne-to-One AdvisoryCONSILIVMPrivate AI counsel, built around your real decisions.90 daysOne executive
- IIIOn-Site ImmersionsIN SITVExecutive training at your assets, built on your data.2–5 daysLeadership teams up to 30
- VSector IntelligenceSPECVLADecision-grade analysis of AI in your sector.From 72 hoursExecutive teams
Begin
A sector briefing for your leadership, within seventy-two hours.
Tell us about your assets and the decision in front of you. A principal replies within one business day.