Researchers have developed a new irrigation system that combines doorbell cameras and artificial intelligence to make lawn irrigation more efficient.
Their recently published paper details this sustainable irrigation system, dubbed ERIC by the researchers.
The ERIC system makes home irrigation smarter, leading to increased efficiency and sustainability. The study shows that ERIC can help users save up to $29 per month in utility costs and as much as 9,000 gallons of water per month for a single residential home, demonstrating significant potential for water conservation.
“We built ERIC with two key components, an existing doorbell camera installed at the residential home and a low-cost irrigation smart controller,” says Tian Liu, a PhD student in the computer science and engineering department at Texas A&M University.
“ERIC analyzes the camera footage using machine learning models to determine how much rain has fallen and automatically adjusts irrigation accordingly.”
The irrigation system analyzes rainfall estimation and then automatically adjusts the irrigation schedule and duration for a property. This more precise irrigation schedule cuts down on excess water use and costs.
This novel approach addresses a long-standing problem in water conservation: inaccurate rainfall data. Most commercial systems rely on imprecise data from weather stations or simple rain sensors that detect the occurrence of rain but not the amount. By contrast, ERIC delivers hyper-local and accurate rainfall estimation using AI-powered video analysis—helping homeowners save water, money, and time.
“Our goal was to make home irrigation both smarter and more sustainable,” says Liu. “And we’ve shown that even affordable, readily available hardware like doorbell cameras can be repurposed to achieve that.”
Radu Stoleru and Charles Swanson, coauthors of the paper, collaborated on a Texas A&M Water Seed Grant Initiative project in 2017 to develop enhanced water efficiency technologies to improve the WaterMyYard program. The program, founded by Swanson and Guy Fipps in 2012, aims to simplify lawn watering for homeowners by providing them with weekly watering recommendations via email, text, or mobile app.
Swanson, a Texas A&M AgriLife Extension Program Specialist, and Stoleru, a professor in the computer science and engineering department, set out to solve several challenges—including the use of data from local weather stations—for the WaterMyYard program, leading to the development of the ERIC irrigation system.
The ERIC system can also solve the challenge of users needing to manually adjust their sprinkler schedules based on weekly rainfall data. Thanks to a type of AI called machine learning, the irrigation system can automatically adjust users’ irrigation systems.
“The biggest challenge we faced was collecting enough diverse rainfall data from real homes to develop and evaluate accurate models,” says Liu. “Due to the scarcity of rainfalls, we spent over two years collecting data, and training and validating models.”
The researchers will continue to collaborate with the Texas A&M AgriLife Extension Service to make the ERIC irrigation system more accessible to the public, deploying and testing the system under the WaterMyYard program.
If successful, ERIC could mark a significant shift in how we manage one of our most precious resources—starting right at the front door.
Source: Texas A&M University
