June 27, 2025

Proactive city leaders use AI to cut water costs

Proactive city leaders use AI to cut water costs

Cutting the cost of water loss

Bartow County Water officials turned to AI-driven leak detection to identify, pinpoint and fix leaks in their water distribution network—helping them cut their annual water costs by $130,000. 

 

The challenge: out of sight, not out of mind

Bartow County Water supplies around 7.5 million gallons of water every day to its 108,000 residents in northwest Georgia, and as the utility acquires over 95% of its water from five external water systems it is strategically imperative that the water is used efficiently and cost-effectively. 

The city’s drinking water distribution network is large and complex, with around 24,000 connections spread across 900 miles of pipes of varying materials and diameters, and like many networks in the US, it suffers from water loss due to leakage. 

Officials knew that the network was losing water to leaks, but as water pipes are buried underground it was unable to detect and locate the leaks that were causing this non-revenue water loss. Resources were being used to fix large leaks and main breaks instead of driving efficiency by proactively preventing them. City leaders decided that they needed to take action. 

“We were mostly fixing leaks instead of finding them, and that was a costly way to address an ongoing problem. We were at the point where our leaks were so bad we just had to fix them without proactively trying to predict and find potential leaks. It was past time to see what we were actually dealing with here.” 

Lamont Kiser, Director of Bartow County Water

The solution: CivilSense™ AI-driven leak detection 

City officials identified CivilSense™ real-time leak detection as the solution that could make their invisible network visible, and they partnered with Oldcastle Infrastructure to address a high-priority area of the network. 

Bartow County Water identified an 80-mile section of the water distribution network that they considered to be at elevated risk of leaks, and the Oldcastle team designed a leak detection project that would provide detailed, specific insight into leak conditions.

Oldcastle experts strategically deployed highly sensitive acoustic sensors at almost 800 different points across the network, gathering detailed acoustic data that was fed into a proprietary AI. 

The AI, powered by FIDO Tech, analyzed the data and used comparative analysis against its curated library of more than 2.3 million acoustic signatures to detect, locate and assess the size of each leak. 

Using this analysis, field teams then redeployed sensors upstream and downstream of each of the identified leak locations and used cloud computing to pinpoint the location of each leak via a technique known as correlation. 

 

The outcome: reduced water cost

CivilSense™ detected nine previously unknown leaks in the section of network, including four large leaks that were discharging up to 20 gallons per minute and which represented a high risk of becoming a costly and disruptive main break. 

The leaks were discovered across a range of network infrastructure assets, including service lines, mains and hydrants. 

In total, CivilSense™ discovered leaks that were losing a combined volume of 83 gallons per minute, which equates to around 43M gallons of water every year. 

By addressing these leaks, Bartow County Water achieved a 17% reduction in non-revenue water that translates to a saving of $130,000 in variable production cost every year. Given the city’s reliance on external water sources, this reduced water cost and increased water distribution efficiency improves not only its sustainability and water resilience, but also its finances. 

The city has subsequently agreed a three-year follow-on partnership for CivilSense™ to analyze additional sections of the network to drive further water cost efficiencies. 

 

Learn more