In This IssueNew This WeekTechnologyTransportation

AI in Infrastructure in 2020

How changing trends in artificial intelligence technology are shaping the world of infrastructure

By Dr. Benjamin Schmidt

Artificial intelligence (AI) as a concept has been around for quite some time now, starting with research that began in the 1980s into how computers might mimic human-like behaviors.

But today, with the continuing rapid advancement of computational power, we’re starting to see AI implemented in industries that were not typical targets for adoption, like tech companies and financial services. From shopping and entertainment to automobiles and healthcare, AI is spreading quickly.

It’s not a disease, though. Quite the opposite. AI is simply the application of advanced computing algorithms to new and novel problem areas. AI not only includes machine learning where we literally teach a computer program something that the original writer of that program did not know (aka, the machine learns); it also includes complex optimization problems using equations, like choosing the best travel path.

With AI, we are now able to develop solutions for critical issues that we couldn’t have otherwise. So how do we see AI expanding and what are its benefits?

Within the past few years, AI start-ups – the usual bellwethers of technological progress – have popped up with intentions of improving and better maintaining our aging infrastructure. From construction
and traffic, to road and sewage maintenance, AI serves as a tool to make everything run more smoothly.

Connecting Construction Workers

“Today we use data to predict where there might be big problems,” explained Terrence Clark, co-founder of Buildstream, a Brooklyn-based construction technology company. is startup utilizes machine learning to understand how a contractor’s work performance can improve in real-time.

Introducing AI into construction allows workers to reduce the amount of time spent on paperwork and instead, focus on completing the task at hand.

Optimizing Traffic Management

One intelligent traffic systems company developed an algorithm to optimize traffic lights in real-time. They discovered this
system not only improved time-savings, but also had a positive environmental impact, with air quality improving due to reduced emissions.

By having efficient traffic control, there was a 30 to 40 percent decrease in stops. This drop spared roads and tires from wear, resulting in better traffic patterns and money savings for municipalities and drivers.

Simplifying Road Assessments

At my company, we’ve discovered that a major value of AI
is objectivity. Currently, infrastructure maintenance is handled visually, and although workers conduct a thorough examination, these assessments are still subjective. Subjectivity leaves more room for error and overall data inconsistency resulting in poorer decisions.

My team uses AI to objectively assess roads. By using a smartphone camera, drivers press record and collect video data from the entire road network. This video is broken down into images which are processed by an AI model trained to identify and rate the severity of road distresses.

Similar to the way we use AI to analyze video data, another startup has sought to use this method for our underwater drainage system.

Saving Sewage Maintenance

Another tech company, Hades, developed a way to use AI to transform how our sewage systems operate. Normally, the sewage maintenance process involves a worker inspecting the area and video recording the state of the infrastructure. Often, defects are missed with human error.

This doesn’t mean that sewage experts are not needed. Machines can outperform humans with repetitive tasks, but humans are much better at complex decision-making.

With AI, sewage experts can supervise the analytics and focus on devising a maintenance plan.


In the end, AI is trending because of its practicality. Across the many facets of infrastructure, we are seeing just the beginning for AI. While this list is in no way comprehensive, it does offer some insights into the breadth of challenges currently being tackled with AI systems.

Artificial Intelligence has, for the most part, been about improving our society, whether it’s finding cures to diseases, redefining voice assistants or simplifying infrastructure maintenance. That trend continues.