
AI robots that can scale vertical walls, crawl along the underside of pipes, and see through structures with infrared cameras, are being used to inspect everything from U.S. Navy warships and Air Force missile silos, to oil and gas storage tanks, bridges and dams. Along with drones, robotic dogs and sensors, companies are offering an all-star team of high-tech gadgets that will elevate the inspection of critical pieces of infrastructure to reduce the number of collapses, explosions or malfunctions that could cost human lives and millions of dollars.
AI robotic crawlers, created by Gecko Robotics Inc., are helping the Defense Department, oil refineries and governmental agencies get a clearer picture of the integrity of their structures and vessels, allowing those entities to be proactive with repairs and maintenance. The crawlers use magnets to cling, like a Gecko, but resemble something out a science-fiction movie, like The Matrix.
Gecko Robotics is one of several companies specializing in AI-powered predictive maintenance to identify potential structural weaknesses, going beyond the capabilities of current monitoring methods, including maintenance workers suspended by harness and using a hammer to tap on walls to get acoustic feedback, sometimes referred to as “Joe on a rope.”
“Our world is old, and it’s crumbling; we rely on power plants, manufacturing sites, infrastructure and even military equipment that is 50, 60, 70, 80 years old, and the way that we insure that those things are going to be there for us are archaic, and that’s why we make Gecko, to insure that we can protect the things that we use and rely on every day,” said Jake Loosararian, Co-Founder and CEO of Gecko Robotics.
The American Society of Civil Engineers publishes a report card every four years assessing America’s infrastructure. The report card gives a grade from A through F on 18 different categories such as rail, aviation, bridges and roads. The last report card was published in 2021 and it gave our infrastructure an overall grade of C-minus, which was up from a D-plus in the previous report card published in 2017. Of the approximately 617,000 bridges in the U.S., four out of ten are at least 50 years old, and 46,154 bridges are considered to be in poor condition, according to ASCE.
But those numbers may be even higher, as human error may miss some detections. “Traditional inspection methods are resource intensive, dangerous and provide limited actionable data, leaving decision makers with insufficient data to make repairs, set maintenance budgets and confidently avoid failure,” Mr. Loosararian said.
The March 26, 2024 collapse of the Key Bridge in Baltimore, MD, which claimed the lives of six construction workers, has highlighted the critical need to upgrade the nation’s infrastructure to prevent such tragedies, and it all starts with better actionable data, Mr. Loosararian said.
He became interested in bringing to market a better inspection method after he visited a power plant in Pennsylvania during his college years. At the power plant, he saw up close a massive boiler that converted water to steam. A worker there told him that the boiler was often off line due to pressure tube explosions, and every day it was offline, there was a $2 million loss. The worker also said that his best friend fell and died a year prior while inspecting the boiler. Mr. Loosararian said he became obsessed with finding a way to use robotics to inspect structures. His company was launched in 2016 and currently inspects tens of thousands of critical pieces of infrastructure.
Through the company’s operating platform, Cantilever, clients supply blueprints, maps, models or digital twins of their structures to create a 3D image. Gecko Robotics then conducts a site visit, with drones, robotic dogs, and then their crawlers, to gather more information on the state of the structure. Those AI robotics fill out the details, down to rust, cracks, dents, leaks, and other wear-and-tear. What is captured is full coverage data, intuitive models and sophisticated digital maps, that are used to generate maintenance planning. The data is so detailed that it can offer the client, as the company has done on numerous occasions, advice on how to improve performance. “It opens up a new way of making decisions with data layers they’ve never seen before,” Mr. Loosararian said.
Through predictive AI modeling, and with fixed sensors on the structures, clients can tap into streaming information about the status of the structure and whether it is operating at an optimal level. This new data can significantly extend the useful life of a structure, generating maintenance schedules even five and ten years down the line.
At an industrial facility in Georgia, Gecko Robotics says it has extended the useful life of the infrastructure by 10 years, representing a $105 million value.
The company Skydio uses autonomous drones to carry out inspections at power plants to stay a step ahead of outages and failures. The drones are able to gather imagery from vantage points that would be impossible for a human to gather on foot. Their drones can capture sub-millimeter-level imagery to reveal the slightest compromises in structures, including corrosion.
“From bridges, to refineries, to construction sites, a Skydio aerial inspection program modernizes any inspection and maintenance plan with automation powered efficiencies, best-in-class sensors, and lower risks for workers,” according to a company spokesperson.
And AI-powered drones promise to revolutionize the landscape of the inspection process with regard to the Navy’s fleet, said Jake Johnson, Enterprise Account Manager at Skydio.
“Utilizing a drone system that is able to create that digital twin, create that log of what the current status of the boat is now instead of flying out multiple inspection teams, people flying across the globe, people doing it again and doing it again and doing it again; in only minutes we’re going to be able to fly a drone system, potentially use third party software to analyze the boat, and see what has changed, what has actually passed the inspection and what has not,” Mr. Johnson said.