Artificial intelligence (AI) is deeply woven into the fabric of our daily lives in so many ways that we can take for granted that it’s there. Spam filters. Smart phones with their personal assistants Alexa and Siri. Autonomously-powered vehicles with predictive capabilities. The cloud — and not those fluffy, cumulous ones floating above us.
Hollywood has glorified AI in films and television for years, dating back as far as 1927 and the film Metropolis, in which AI takes the form of a humanoid robot intent on taking over a mega-city by inciting chaos; on the TV side we had The Jetsons. In these instances, they offered exaggerated, make-believe glimpses of how the technology can be used. In everyday life, AI technology may not be as exciting and epic; however, its impact on the construction industry has been exactly that as it has brought an added level of efficiency, accuracy and safety to the construction landscape.
For the world of trenchless technology, we used to speak of AI technology in the future tense. No more. The future is now for AI use in the trenchless market, as it continues to integrate into several trenchless application segments.
Most notably, AI has had an immediate and visible impact in CCTV pipe inspection and has quickly gained traction as an important technology tool in the software toolbox in the last few years. Pipe inspection software companies have been in the research and development stages for AI for many years, working to perfect its inclusion in the inspection process through images and accurate coding. Companies such as WinCan, SewerAI, and Fracta, to name a few, have rolled out AI software packages to enhance technology already utilized to more accurately gather and assess the condition of the underground pipes.
“AI is the ability to learn and solve problems the way the human mind does,” says WinCan general manager Mike Russin. “In relation to condition assessment, AI is already improving the speed and accuracy as to which data is collected and reviewed. Now as AI evolves into work management and financial planning on the data, we see tremendous benefits for years to come in the trenchless industry.”
AI technology for pipe inspection has garnered the most attention but AI has been incorporated in other trenchless applications including new installations segments such as microtunneling, horizontal directional drilling (HDD), pipe jacking and auger boring, to name a few. AI use here has provided for more accurate data reporting and logging, safety features and guidance/steering.
“AI definitely has a place in new installation projects and aid in the production of doing an auger bore, microtunnel or HDD bore,” says Scott Fisher, international sales manager and technical advisor at Barbco, a worldwide maker of auger boring machines, directional drills, guided boring machines and advanced tunneling equipment. “AI is not going to do the job for you but it will make it easier, safer, more efficient and cost-effective. It’s there to assist us in completing jobs and keep our operators safe.”
How AI Works in Trenchless
Let’s take a look at how AI is being used on jobsites today, starting with CCTV pipe inspection. In general terms, AI will offer the ability to auto-recognize defects from CCTV inspections and produce an accurate grade for the condition of said pipe. Its proponents say AI has the potential to produce this data more quickly and accurately than the human eye does right now.
And part of that efficiency and accuracy is having NASSCO certification of the process. NASSCO’s Pipeline Assessment Certification Program — better known as PACP — has been an integral part of the pipe inspection process for more than 15 years. Using its 1-5 coding system (with 5 being the most severe condition), PACP has given a universal voice to identifying and recording pipe defects, deformities, deterioration and other issues. Any use of AI technology will need to be able to replicate this very specific coding system.
“Opportunities right now [for AI] are in the visual aspects of the data,” Russin says.
There is a gamut of different equipment available to collect data from the pipes — CCTV to drones, push cameras, 360 cameras, HD scanning. As cities become inundated with the amount of data they are collecting and the time it takes to assess said data, AI can help them get out from underneath it.
“Cities are getting inundated with data causing choke points that delay condition assessment workflows,” Russin says. “The opportunity for AI right now is to lessen the choke points and start processing data more efficiently. With these efficiency gains, end users are able quickly build accurate data models for efficient review of their overall wastewater network”
WinCan launched its Sewermatics program in 2021 leveraging AI to provides users with AI-assisted defect coding ,data conversion, data visualization all hosted in an intuitive cloud platform. All of these functions give users a more accurate condition assessment of the infrastructure, strengthening decision-makers information on which to select and budget for the appropriate rehab/repair program.
SewerAI was founded three years ago in Walnut Creek, California, and provides professional services that use AI tools — a cloud-based platform built for data management. Customers upload the data and SewerAI executes the AutoCode process for pipe assessment. SewerAI director of business development Eric Sullivan sees the impact AI technology is having today and is excited for what lies ahead.
“AI tools generate information for three categories: descriptive, predictive and prescriptive,” Sullivan says. “Ours focuses on descriptive and using AI to identify and describe the conditions that are in the pipe. We are starting to leverage that data through other AI processes where they can yield predictions that humans would not be able to calculate.
“We’re not relying on the AI tools to tell us to do or not to anything,” he says. “We’re using the tool to help us more reliably and efficiently identify what is or is not in the pipe, manhole or sewer.”
Sullivan also sees AI impacting the workforce of operators — in a good way. He says the CCTV industry is also struggling find good operators who can film and analyze the pipe data. “While they still need NASSCO training, operators [with AI] do not themselves need to be experts in analyzing the condition of a pipe,” he says. “AI is a pretty powerful tool to all of our industry to be more inclusive of newer people who might not otherwise be suited for the role.
“We’re seeing a pretty major impact on contractors being able to complete projects much quicker and more efficiently and at a higher level of consistency in terms of the final product,” he says. “The benefit to the lining companies is that they are better able to get the information they need to plan for various types of installs, especially identifying problematic conditions that might pose issues for certain CIPP lining installs.”
New Installation AI Opportunities
While AI may be more visible on the pipe inspection side, the technology is being used in new installation applications, primarily in areas of data logging, crew safety and guidance. In smaller ways, AI technology has been incorporated into applications such as microtunneling, pilot-tube guided boring, HDD, pipe jacking and auger boring.
“AI is a way to enhance quality, productivity and provide a level on internal inspection,” says Akkerman vice president and chief revenue officer Jason Holden. “AI has been used longer than we have been thinking that it was ‘a thing.’ Right now, the parameters are contained within the guidance system and just does the reporting,” he says. “Nothing is linked to machine control because the geotechnical and ground conditions that the machines go through can’t be fully known. That’s always the underlying factor.
“We see AI in the control systems, behind the scenes for warnings and safety checks and data reporting,” Holden says.
One area Akkerman is testing AI on is with autonomous muck hauling, with the end-game being able to drive the hauling system between the launch shaft and the TBM system to retrieve the processed muck from the reception shaft. “We’re testing this in-house to see other ways we can use the AI technology in other systems,” Holden says.
Fisher is a proponent of AI technology and its use in making the boring operations safer, even though some old-school thinking contractors have been slow to come around to that way of thinking. He sees the technology as being able to reduce labor costs vs. getting rid of labor costs. “The efficiency and accuracy are where it’s going to pay off,” he says. “The straighter you can put in a pilot hole and follow that pilot hole, the more accurate your job will be.”
One area that Fisher says AI is used is in Tilt Anticipation Shutdowns for Barbco auger boring machines. In general terms, if the machine senses it will exceed previously set parameters of grade, the machine will shut down and disengage the clutch to aid in preventing the machine from flipping over. But this AI-aided safety enhancement isn’t a favorite for all operators, as the machine would be stopped multiple times as sensors react to how the machine is moving.
“Getting older contractors on board with this was a challenge,” Fisher says. “They didn’t like the stoppage and they didn’t trust the buttons to be fast enough for a human reaction [, if needed]. They just wanted to run the machines. I was used to running the old machines too, but I could see the benefit [of the technology] because if I’m on a tough job I don’t want to put my life in danger. If there is something available, I want to use it.”
Human vs. Machine
Will AI ever reach the point of replacing humans performing the work? Not likely. No matter how much or how little AI becomes integrated into a trenchless application/system, it won’t take the place of a human, our panelists say.
“It’s impossible to replace the experience of an operator,” says Holden. “I don’t think AI will ever be in a position to replace them.”
Holden says for installation applications there is a safety factor to consider. “In the confined spaces we work in, it can be dangerous and require reactions and quick thinking. That’s built on years of experience. Fully autonomous tunneling is years and years away. There is the brain power and it can be done but the cost is high, as are the risk and exposure.”
Fisher agrees that operator-less boring and drilling is not a consideration at this point, primarily due to safety considerations. “If you’re thinking that you want to put a machine without humans operating on the job and said do this bore, there are so many factors against it,” he says. “Auger boring, for example, may seem very simple on the outside but it’s very complex on the inside. There are so many variations in ground conditions, weather, groundwater, soil hardness and to have a machine that’s able to deal with all those factors isn’t possible. I could take all the people I’ve been associated with in the last 35 years and take all their knowledge and it’s still not enough. How could you program a machine to do all that?
Regarding pipe inspection, Russin and Sullivan say AI technology enhances what is already available. “AI will not eliminate jobs,” Russin says. “It will produce accurate results more quickly while supporting field and office personnel, not replace them. The AI models still rely on human supervision so as training continues the models will improve and become more beneficial.”
“It’s not replacing anything right now, other than traditional workflows,” Sullivan notes. “In the next few years, you’ll see AI more commonly used than not. The technology has progressed very quickly. The market adoption is more a question of educating the stakeholders.”
In the future, AI will be used in more ways as the technology gets better and smarter. But that doesn’t mean humans will not be needed. “Machines aren’t smart because they are machines,” Fisher says. “They are smart because humans made them smart and there will still be humans involved in AI use.”