Experiential Learning and Automation of Insect Pest Detection

Goals / Objectives
Employ modern acoustic sensor systems and and construct artificial intelligence (AI) deep learning algorithms to develop improved methods to detect and inform control of hidden infestations of invasive insects, and train students of Florida A&M university, an HBCU, to apply such methods in service as researchers and leaders in a highly skilled upcoming generation of a diverse and inclusive USDA workforce.

Project Methods
Over the last 30 years, acoustic methods have been developed at the USDA Center for Medical, Agricultural, and Veterinary Entomology (CMAVE) to detect hidden infestations of insect pests in trees, soil, and stored products, and develop tools to help protect against agricultural losses and environmental harm. Important progress has been made in distinguishing insect sounds from background noise and in distinguishing pest insect species from each other. One of the most promising tools is the use of deep learning algorithms that will be developed and calibrated under this research. Students will participate in the data collection process and in inspections to confirm infestation. They will participate also in the calibration process that will be led by the ARS PI. The FAMU PIs will provide logistic support. Manuscripts on the results of the study will be written and published by the students, FAMU PIs and ARS PI.