Students at Spelman College are harnessing artificial intelligence to help people better understand and care for plants through a new technology they call PlantGPT. The innovative tool, developed by a team of student researchers, uses sensors to collect environmental data from plants and translate those readings into real-time care guidance. By analyzing the data through an AI-driven system, PlantGPT can recommend when plants need water, more sunlight, or other adjustments, offering users an accessible way to monitor plant health and improve plant care.
The project is being developed through the Arthur M. Blank Innovation Lab at Spelman, where a team of student innovators has contributed to various stages of the technology’s development. The group includes students Joy Rutledge, Temple Dees, Jessica Obi, and Devyn Washington, and alumna Grace Burch, who have worked collaboratively to design, test, and refine the system.
“PlantGPT is a way to be able to talk to your plants,” Rutledge said in an Instagram video produced by the college. “So when your plant is dying, you need to know what type of information the plant needs or resources that the plant needs, such as water, sunlight, or nutrients.”
Using sensors placed near plants, the system gathers environmental readings to monitor growing conditions, including soil moisture, humidity, temperature, and light exposure. That data is transmitted to an artificial intelligence model that interprets the readings and generates recommendations for plant care.
“So we’ve basically created sensors and connected it to an AI,” Dees explained. “You’re able to get this real-time data from your plant so that when it notifies you, you can adjust the conditions of your plant.”
The current team is continuing work that began with Burch, who graduated from Spelman with a degree in biology. Her early research helped develop the project’s initial prototype as part of experimentation within the Innovation Lab.
Since its early stages, PlantGPT has undergone several technical upgrades to improve the system’s capabilities. An early version of the project relied on Arduino-based hardware, but the design has since evolved as the team refined both the sensors and the artificial intelligence model. The prototype is currently being tested on houseplants while researchers continue improving its functionality.
“My part of the project was switching the AI model to TinyLlama, which is local and doesn’t need the internet, and adding more features like written responses,” Dees said. “We’re also working on adding a vocal response feature.”
While still in development, PlantGPT has already earned recognition and early funding. The project received a $1,000 award at the Center for Black Entrepreneurship’s New Venture Competition in 2025 and won top honors at the American Association for the Advancement of Science Making & Innovation Competition in Washington, D.C.
