Redefining AI for Agriculture and Forestry Systems through Domain-Adaptive, Physics-Aware Co-Design
 

Abstract

Many challenges with artificial intelligence and machine learning (AI/ML) for Agriculture and Forestry arise from the physical and environmental conditions in which these systems operate. Extreme temperature, humidity, wind, atmospheric effects, and geolocation introduce substantial variability and uncertainty that degrade data quality and limit the reliability of data-driven, physics-agnostic AI approaches. While canonical AI methods often assume relatively constrained data conditions, real-world agricultural and forestry environments demand resilient systems capable of operating across large spectral, temporal, and spatial variations. 

This talk introduces a new paradigm for domain-adaptive AI that explicitly integrates AI/ML with optical physics. By co-designing learning algorithms and sensing modalities, this framework models the physical processes underlying image and signal formation to improve robustness, generalization, and trustworthiness. I will highlight recent advances at the intersection of computer vision, signal and image processing, biometrics, and optical physics that address challenges such as spectral variability, limited labeled data, and spatially and temporally varying atmospheric turbulence. 

These methods are motivated and validated through applications in precision agriculture and precision livestock systems. I will conclude by outlining emerging research directions and a long-term vision for building sustainable, interdisciplinary research programs in AI for Agriculture and Forestry, where bridging AI/ML and optical physics to enable reliable decision-making in complex, real-world environments directly impact national food security and ecosystem resilience.

Biography

Benjamin Riggan, Associate Professor in the Department of Electrical and Computer Engineering at the University of Nebraska–Lincoln (UNL), has research that is focused on domain adaptation, optical physics-informed sensing, image and signal processing, and biometrics, with applications in challenging real-world environments where data quality and variability impede traditional AI systems.

Riggan received his BS in computer engineering, MS in electrical engineering, and PhD in electrical engineering from North Carolina State University in 2009, 2011, and 2014, respectively. Prior to joining UNL in 2019, he was a Research Scientist and Postdoctoral Fellow at the US Army Research Laboratory’s Image Processing and Networked Sensing branches, where he conducted research on long-range and cross-spectrum recognition.

His work has been supported by agencies including USDA, IARPA, Army Research Laboratory, National Strategic Research Institute, and he currently serves as a Senior Editor for the IEEE Transactions on Aerospace and Electronic Systems. He has published extensively in top peer-reviewed venues, received multiple Best Paper Awards, and contributes leadership to the broader scientific community through conference organization and editorial service.