AI Detects Sewer Line Defects: CU Denver Engineers Develop Innovative Tool
Denver Sewer System Gets a Smart Upgrade Thanks to CU Denver Innovation
Table of Contents
Denver, CO – Keeping Denver’s 1,500 miles of sewer lines flowing smoothly is a monumental task. Breaks can cause major disruptions, shutting down businesses and even forcing the state capitol to close its doors, as happened in August 2024. Now, a CU Denver graduate student and his professor are using artificial intelligence (AI) to revolutionize how the city inspects and maintains its vast underground network.
Brian Zavareh, a doctoral student in CU Denver’s College of Engineering, Design, and Computing, witnessed firsthand the time-consuming process of sewer inspection during a 2019 internship with Denver’s wastewater IT department. Crews spent countless hours reviewing video footage from inside pipelines, searching for over 250 different types of pipe failures.
“I thought I could help by automating some of the process,” Zavareh saeid. “I offered to develop an ML [machine learning] model to help the operators, and it worked.”
Zavareh’s model can process five hours of video footage and identify defects with over 85% accuracy in less than 30 minutes, a significant improvement over manual review.
Faculty-Student Mentorship Fuels Success
zavareh credits his success to the guidance and support of his mentor, CU Denver computer science and engineering professor Farnoush Banaei-Kashani.
“Brian’s internship gave him the chance to identify a real-life problem,and he came up with an idea to solve it,” Banaei-Kashani said. “That’s what we engineers do.”
Zavareh’s journey began with a simple desire to become a software developer. However, a class with Banaei-Kashani sparked a passion for data science and its potential to address large-scale problems.
“Your relationship with your faculty is instrumental to success,” zavareh said. “He’s been nothing but enthusiastic and supportive of my ideas. That’s why we are here today.”
From Classroom to Start-Up
The professor-student duo has taken their innovation beyond the classroom,launching InfraSmart Solutions,a start-up company through CU Denver’s Smart Futures Lab. The lab provides an incubator and accelerator program for budding entrepreneurs.
“Right now,we are at a base model,” Zavareh said. “I’m really excited about the potential once implemented. I can’t wait to see it grow.”
InfraSmart Solutions is currently in talks with the city of Denver about a pilot program and is exploring potential projects with two other municipalities.
A Future of Smarter Infrastructure
The team is continuously refining their AI tool, exploring ways to enhance its capabilities to manage work crew dispatch and multiple projects simultaneously. this approach has the potential to be applied to other types of infrastructure, including bridges and airports.
“AI is an amazing technology that keeps evolving,” Banaei-Kashani said. “We’re working hard at CU Denver to innovate and apply what we are seeing in the lab to industry and business. We’re training our future scientists and engineers. It is one of the most exciting times to be in this field.”
Denver Sewers Get Smart Upgrade Thanks to CU Denver Innovation: An Interview with Brian Zavareh and Professor farnoush Banaei-kashani
NewsDirectory3: Denver’s vast sewer system, spanning 1,500 miles, faces the constant threat of breaks and disruptions. Now, a CU denver graduate student and professor are using artificial intelligence to revolutionize how the city inspects and maintains its underground network. We spoke with Brian Zavareh, a doctoral student in CU Denver’s Collage of engineering, Design, and Computing, and his mentor, Professor farnoush Banaei-Kashani, to learn more about this innovative solution.
NewsDirectory3: Brian, your AI model can analyze hours of video footage and identify pipe defects with impressive accuracy. What inspired you to develop this technology?
Brian Zavareh: During a 2019 internship with Denver’s wastewater IT department, I saw firsthand the time-consuming nature of sewer inspection. Crews spent countless hours reviewing video footage, looking for over 250 different types of pipe failures. I thought I could help automate this process and make it more efficient.
NewsDirectory3: Your model can process five hours of video footage and identify defects with over 85% accuracy in less than 30 minutes. How does this compare to manual inspection?
Brian Zavareh: This represents a important betterment over manual review,which can take significantly longer and is prone to human error.
NewsDirectory3: Professor Banaei-Kashani, how has Brian’s passion for data science led to this impactful innovation?
Professor Farnoush Banaei-Kashani: Brian’s internship provided him with a real-world problem, and he came up with an ingenious solution. His work exemplifies the power of data science to address crucial infrastructure challenges.
NewsDirectory3: You have co-founded InfraSmart Solutions, a start-up company aimed at commercializing this technology. Can you tell us about your vision for the future?
Brian Zavareh: We’re currently in talks with the city of Denver for a pilot program and exploring potential projects with two other municipalities. Our long-term vision is to see our AI-powered solution implemented across various infrastructures, including bridges and airports.
NewsDirectory3: Professor Banaei-Kashani, what are your thoughts on the role of CU Denver in fostering such innovative solutions?
Professor Farnoush Banaei-Kashani: CU Denver is dedicated to training the next generation of scientists and engineers. We strive to bridge the gap between academia and industry, ensuring our research translates into real-world applications. The success of Infrastructure Solutions exemplifies this commitment.
NewsDirectory3: Thank you both for your time and valuable insights. Your work promises to revolutionize how we manage and maintain critical infrastructure for generations to come.