Verdis wins Rep. Keith Self’s 2025 Congressional App Challenge in Texas’s Third District

When asked what inspired the creation of Verdis, the students said, “Around our hometown of Prosper, Texas, we have many huge cropfields where we see farmers working tirelessly throughout the year on tractors and combines. One of our group members’ families is from a farming background overseas, so we had first-hand experience with the tremendous amount of effort that goes into the cultivation and mass production of crops. We also wanted to do a project that integrated mechanical, electrical, and software engineering, since each of us has experience in one of these areas, and we wanted to combine our strengths.

“After we decided to look for ways to reduce this effort in an inexpensive and repeatable way, we landed on monitoring the health of crops, which we learned is an important task for keeping yield as high as possible for farmers. Diseases spreading that are unknown to farmers can be devastating, and minimizing diseased crops can help control the total spread. According to Sustainable Agriculture Research and Education, up to $50 can be lost in each acre of farmland per year, which adds up around the globe to billions of dollars, which could be used to feed an innumerable number of people.

“The most effective way to monitor crop health is through aerial scans, specifically with an autonomous drone. A drone has an incredible range, from skimming the ground to 400 feet in the air (a maximum altitude regulation from the Federal Aviation Administration for civilian drones). This allows it to take images from altitude and even descend to unclear spots for closer inspection. However, seeing that drones capable of this cost several thousand dollars inspired us to build our own, combining it with an application in order to achieve comparable functionality at a lower cost while also accurately addressing challenges that farmers face. After deciding the best way to tackle this problem, we started designing the vehicle and programming machine learning algorithms using existing datasets, eventually leading to our final product.”