“I grew up in India where water was a dear resource, something that had to be rationed. Over half the water was wasted because of bad pipes,” says Sanjay Sarma, an associate professor of mechanical engineering.

As Sarma observes, the world’s water distribution systems are notoriously inefficient. By some measures, in the most advanced systems, like those in the U.S., nearly 30 percent of the water never reaches its destination. In developing countries, the problem is even more acute, losing as much as 50 percent of the water. And in the arid regions of the world, where fossil fuels are burned for desalination, CO2 is being unnecessarily released into the atmosphere, as well.

The problem lies in the difficulty of finding cracks and vulnerabilities in the water system. Traditionally, hydrophones have been used to listen for water pressure waves, but the process is imprecise. Moreover, hydrophones can usually only detect large leaks.

So Sarma wondered, what do small leaks sound like? To find out, he and his colleagues are developing an in-pipe mobile acoustic sensor that can detect and localize problems in the water network. In other words, they’re building a robot that “listens” for the sound of small leaks and then reports their location.

It’s an ambitious project with significant challenges. For one, Sarma says, the robot is looking for the “whisper” of the leak, just a trickle amid rushing water. Further, the robot has to be able to communicate from deep within the ground. “So once the robot hears the whisper, who hears the robot? We can’t collect the robot at the end of its mission, because we could lose it in the pipes, along with valuable information,” says Sarma. So the professor is hoping to apply his world-renowned expertise in radio frequency to build a specialized communications system. And finally, the robot has to be somewhat self-propelled, able to navigate a tight network of underground pipes because, as Sarma quips, “I haven’t yet found a graduate student small enough to fit.”

If the robot works, Sarma is poised to solve a costly problem that has long perplexed local governments. Once a leak is suspected, they must dig up meters and meters of pipeline in hopes of finding and correcting the faulty pipe. And while his project is currently in the instrumentation phase, Sarma has already had some promising results that have attracted collaborators from around the world.

Sarma’s group is currently doing some simulations that will help build a prototype by laying pipe underground and testing their communications, sensing, and mobility instruments in a variety of conditions. “I’m optimistic that this will work; but for now, we’re literally laying the groundwork,” Sarma says. “This is just the beginning.”