UNIVERSITY PARK, Pa. — Whether identifying a plant or animal disease, interpreting a soil test, or finding educational resources, users have turned to Tilva more than 40,000 times since the Penn State Extension artificial-intelligence (AI) tool launched in January. New features introduced in recent months are designed to help users continue those conversations in English or Spanish and connect with additional sources of trusted, research-based guidance.
Tilva is an AI-powered tool that provides around-the-clock, research-based answers tailored to Pennsylvania agriculture for farmers, agribusiness professionals, educators and others.
“Tilva also provides answers well beyond traditional agricultural production issues, supporting Extension’s broad mission to serve the citizens of the commonwealth,” said Troy Ott, Peter and Ann Tombros Dean of the College of Agricultural Sciences. “From youth development, natural resource management, diet, health and many other topics, Tilva provides answers that Pennsylvanians can trust.”
While most questions come from Pennsylvania users, people in 152 countries have submitted questions through the platform.
“While Tilva is being designed to ensure it’s ideal for Pennsylvania, it’s been encouraging to see people from around the world engage with extension resources through Tilva,” said Jeffrey Hyde, director of Extension and associate dean in the college.
Tilva allows users to ask questions or upload photos. The system draws from decades of science-based extension expertise to deliver reliable information on topics such as crop production, pest management, horticulture, food systems and natural resources.
Justin Clapper, assistant director of development and sponsorship for Pennsylvania Farm Bureau and a board member of the Penn State Ag Council, has used Tilva to explore questions about production agriculture, animal health and plants. This year, he planted new varieties in his garden and used Tilva to learn more about the plants.
“I think the most useful aspect is that unlike other AI models that pull random information from the internet, Tilva cites reviewed information and articles from Penn State Extension,” Clapper said. “It also recommends other extension articles and resources for you to pursue and dig deeper into a topic if needed. You can trust the information is accurate.”
Since its launch, project leaders have introduced several updates to help users explore topics in greater depth and connect with additional extension resources.
One new feature suggests follow-up questions after a response is generated, encouraging users to continue exploring a topic.
Tilva also now includes an option for users to subscribe to Penn State Extension newsletters related to subjects they are researching. Depending on the topic, the system may recommend newsletters covering areas such as agronomy, horticulture, livestock production or other extension program areas.
In addition, the platform can recommend educational resources associated with a user’s question, including publications, workshops, online courses and certificate programs offered through Penn State Extension.
Every response now includes an option to request expert consultation, allowing users to connect with a live human for follow-up information or conversations.
Tilva is powered by PlantVillage technology, developed under the leadership of David Hughes, the Dorothy Foehr Huck and J. Lloyd Huck Chair in Global Food Security in the college. What began as a mobile tool to help farmers diagnose crop problems has grown into a global agricultural AI platform used in dozens of countries.
Behind Tilva’s responses is oversight from more than 260 extension educators across 67 counties.
“Artificial intelligence has tremendous potential, but its value depends on the quality of the information behind it,” Ott said. “Tilva helps extend the reach of Penn State Extension by making research-based knowledge easier to access while keeping our educators at the center of the experience.”
He continued, “Tilva represents the best of Cooperative Extension — a more than 100-year tradition of trusted, science-based information, and a commitment to continually renew and reimagine how Extension can best serve the commonwealth.”
Looking ahead, Clapper said he’s excited about the broader possibilities of AI as a tool to support livestock producers. He has spent nearly 20 years in the beef industry as a cattleman.
“It would be great if, in the future, you could scan an animal and it could give you answers related to its health, such as, ‘This animal has a temperature,’ ‘This animal is pregnant,’ and so on,” he said.
As Tilva continues to evolve, project leaders said future enhancements will focus on making research-based information even easier to access while maintaining the quality and reliability of responses.