Why I Can't Stop Thinking About Sewers, Pipes, and Power Lines

Siemens Energy Accelerates Power Grid Asset Simulation 10,000x Using NVIDIA PhysicsNeMo.

One of the standout themes at New York Climate Week was the sudden popularity of infrastructure and utilities. Conversations with fellow attendees, along with reflections shared online (shoutout to CTVC), all pointed towards one key takeaway: AI and advanced analytics are rapidly transforming these traditionally overlooked sectors into hubs of technological innovation.

Infrastructure and utilities have the the reputation of being dull and bureaucratic. But this outdated perception misses the bigger picture. We are all aware that much of the world’s physical infrastructure is old and inefficient. In the U.S., many bridges were built before our grandparents were even born. Power grids, water systems, transportation networks, and waste management facilities often rely on technology and designs that predate the digital age. At the same time, the workforce maintaining these systems is aging rapidly, creating gaps in expertise and operational capacity that are ready to be addressed.

These converging trends are creating a moment of both great necessity and opportunity. Artificial intelligence and advanced analytics are beginning to modernize operations, optimize efficiency, and extend the life of infrastructure, while public and private capital mobilizes to fund upgrades at scale. Basically, what has long been considered “boring” or “unsexy” is becoming a frontier for technological innovation and investment. The intersection of aging assets, demographic shifts, and AI-driven modernization signals that infrastructure and utilities are not only essential, but, for the first time in many, many years are among the most compelling areas for strategic investment.

The Age and Inefficiency of Physical Infrastructure

Much of the physical infrastructure that underpins modern life is decades old, and in many cases, dangerously outdated. In the U.S., for example, the average bridge is over 45 years old, and 40% of major roads in the country are in poor or mediocre condition. Water systems tell a similar story: thousands of miles of cast-iron and lead pipes, some installed in the early 20th century, continue to transport drinking water to millions of households, often with significant leakage and maintenance issues. The electric grid, too, relies on transmission and distribution networks designed in an era before the digital economy and renewable energy integration. These systems were built to meet the needs of the past, not the complexities of today’s urbanization, climate stress, and energy demand.

While this aging infrastructure presents clear risks, it also creates an extraordinary opportunity. Outdated systems are inherently inefficient (simply because we have literally invented better ways to do things), but this makes them ideal candidates for modernization and technological intervention. Advanced analytics, AI-driven predictive maintenance, and sensor networks can identify vulnerabilities before failures occur, optimize resource allocation, and reduce operational costs. Importantly, public and private capital is becoming increasingly available to fund these upgrades, from infrastructure legislation in the U.S. (sans recent actions from the current administration) to international development programs and private equity initiatives. In other words, these sectors are not just in need of repair, they are a fertile ground for innovation and investment, where improvements can yield outsized social, environmental, and financial returns.

Workforce Challenges and the Silver Tsunami

And if crumbling bridges and aging water mains weren’t enough, there’s another quiet fault line running through the infrastructure sector: its workforce. The people who built and maintain these systems are retiring faster than they can be replaced. In the U.S., more than a quarter of utility workers are over 55, and by 2030, roughly half of the skilled trades workforce could be eligible for retirement. This “Silver Tsunami” represents not just a loss of manpower, but a loss of institutional memory; decades of tacit, analog knowledge about how critical systems actually function.

As these experienced workers retire, the next generation is less inclined to enter slow-moving, bureaucratic industries. That widening skills gap is forcing utilities and governments to modernize, embedding intelligence directly into infrastructure to offset shrinking workforces. Automation, digital twins, and AI-driven asset management can capture operational knowledge, streamline maintenance, and make infrastructure more resilient. In that sense, the labor crunch isa catalyst that could accelerate long-overdue transformation across some of the most essential systems in the world.

AI and the Infrastructure Revolution

The good news is that this transformation is already underway. AI, no longer confined to software and consumer tech, is reshaping how those working in infrastructure and utilities manage the physical world. Industry operators are beginning to deploy AI to monitor assets, predict failures, and optimize performance at a scale that human teams alone never could. What was once reactive (that is, fixing things only after they broke) is becoming predictive and preventative.

Energy companies like Siemens Energy and National Grid ESO are using machine-learning models to forecast demand and detect grid vulnerabilities before outages occur (random aside — Siemens Energy trucks are ALL over Columbia’s campus, and noticing them is what originally inspired me to write this piece). In water management, firms such as Xylem and Suez have introduced AI-powered leak detection and smart metering systems that save millions of gallons each year. Even transportation networks are adopting predictive maintenance tools that cut downtime and emissions across rail and aviation. Together, these advances represent more than incremental efficiency gains, they mark the digitization of one of the world’s most essential sectors.

Closing thoughts

For investors, this convergence of AI and physical infrastructure should be really, really interesting. These are trillion-dollar sectors that are literally undergoing a technological revaluation! The next great infrastructure buildout (which, as an American, I can only fantasize about) seems like it will be a combination of concrete, steel, software, sensors, and embedded learning algorithms.

To wrap things up nicely, here are some companies of various stages doing pretty cool work in bringing AI and data-driven innovation to the systems that keep the world running.

1. SewerAIMaking underground infrastructure management wayyy better through its AI-powered platform, PIONEER™, which automates sewer inspections and defect detection.
2. GE VernovaUsing AI-powered Autonomous Inspection to automate visual inspections of energy infrastructure, speeding up asset monitoring, improving safety, and optimizing performance.
3. Siemens EnergyUsing Senseye Predictive Maintenance and Omnivise Predictive Solutions to enhance asset reliability and performance across power grids and industrial plants.
4. XylemUtilizing AI-powered Xylem Vue and smart sensors to automate leak detection, monitor water quality, and optimize distribution networks.

Finally, an observation I had while researching: because their B2B customer base is largely made up of aging baby boomers who may be wary of all the “AI” hype (which I mean, not completely unjustified, especially in critical industries), infrastructure and utility companies often don’t advertise AI as openly as other industries. Instead, they tend to use terms like “predictive maintenance,” “asset optimization,” or “intelligent monitoring,” highlighting practical benefits rather than the technology itself.