Speaker: Dr. Hsiao-Mei Wu, Associate Professor, Department of Biomechatronics Engineering, NTU

Talk Overview: Developing reliable quantitative methods for early detection of plant stress is a critical challenge in agricultural phenotyping. Traditional methods that rely on fluorescence intensity to estimate photosynthetic efficiency are often subject to noise. In contrast, fluorescence lifetime, which is sensitive to molecular states and environmental conditions, can reveal early physiological changes in plants. This presentation introduces the novel FD-FLI system, which was developed to characterize the photosynthetic efficiency of leaves under abiotic stress. Due to its high sensitivity, the system holds promise for long-term, non-invasive monitoring of plant health in the early stages.

About Dr. Hsiao-Mei Wu
Professor Wu is an Assistant Professor in the Department of Biomechatronics Engineering at NTU. Her research interests include the development of optomechatronic systems for applications in agriculture and biotechnology, including the use of fluorescence lifetime imaging systems to measure the photosynthetic efficiency of leaves and characterize oxygen effects on plant roots. In addition, their research groups combine several AI-based image analysis techniques with optomechatronic systems to automatically monitor cells/plants and further reconstruct the three-dimensional cells/plants.

Speaker: Dr. David Bullock, Professor, Agricultural and Consumer Economics, UIUC

Talk Overview: On-farm precision experimentation (OFPE) is a burgeoning and revolutionary method of generating agronomic data.  The principal technique of OFPE is to design spatial-polygon-based computer files (“shapefiles”) that provide “prescriptions” that farmers can use with variable rate input application equipment combined with GPS to put agronomic field trials “in the ground” while the machine operator simply drives. This allows whole-field “checkerboard” field trials to be implemented by the farmer with relative ease, using equipment and techniques with which many North American farmers are well familiar. The concept of OFPEs arose soon after the commercialization of precision technology in the 1990s.  Over the past decade, the USDA-sponsored Data-Intensive Farm Management (DIFM) project has facilitated the growth of OFPE on farms in dozens of U.S. states, four Canadian provinces, as well as in South America and South Africa.  DIFM has recently developed its publicly-available, user-friendly “cyber infrastructure,” which can be accessed through https://difm.farm.  DIFM and participating farmers and crop consultants will run approximately 100 OFPEs in 2025.  But the difm.farm cyber-infrastructure is scalable, and DIFM envisions its use in thousands of OFPEs, all over the world, in the near future.  University of Illinois Professor David Bullock is DIFM’s Principal Investigator.  The purpose of his presentation is to demonstrate how real-world farmers and farm advisors can use difm.farm to run OFPEs on their own fields and benefit from the data generated.

About Dr. David Bullock
David S. Bullock is a Professor in the Department of Agricultural and Consumer Economics at the University of Illinois.  He studies the economics of agricultural technology and information and has published research on precision agriculture technology since 1998.  He is the Principal Investigator of the USDA-sponsored Data-Intensive Farm Management project, which uses precision agriculture technology to conduct large-scale, on farm agronomic experiments which generate data to aid farmers’ management of nitrogen fertilizer and other inputs.   He teaches graduate courses in microeconomic theory.  He received his Ph.D. from the Department of Economics at the University of Chicago in 1989. 

Speaker: Dr. Shih-Fang Chen, Associate Professor, Department of Biomechatronics Engineering, NTU

Talk Overview:
In crop production, harvesting, and post-harvest handling, traditional methods rely heavily on human labor for monitoring and decision-making. However, the shortage of skilled experts and declining labor availability present challenges. Image-based plant phenotyping offers a transformative solution by automating crop trait identification and documentation, reducing reliance on manual assessment. ntegrating IoT sensors, autonomous vehicles, image processing, spectral sensing, and AI algorithms enables expert-level decision-making with greater efficiency. This presentation explores how these technologies enhance agricultural practices, with case studies in vegetables, fruits, and coffee covering crop monitoring, pest detection, harvest forecasting, quality grading, and flavor prediction. These innovations optimize labor use and provide new insights into food flavor science.

About Dr. Shih-Fang Chen

Dr. Shih-Fang Chen is an Associate Professor in the Department of Biomechatronics Engineering at NTU. She earned her Ph.D. in Agricultural and Biological Engineering from the University of Illinois at Urbana-Champaign. Dr. Chen’s research specializes in the application of image processing and spectral analysis techniques for plant status monitoring and agricultural product quality evaluation. Her work is centered on the advancement of smart agriculture through the integration of Internet of Things (IoT) and Artificial Intelligence (AI) technologies to enhance precision farming methodologies. Her research portfolio encompasses a diverse range of agricultural applications, including plant disease identification, harvest time prediction, agricultural product grading, pesticide residue detection, and flavor prediction systems. She has served as a Guest Editor for Special Issues of Computers and Electronics in Agriculture. She currently holds the position of Associate Editor for Engineering in Agriculture, Environment and Food.

Speaker: Dr. Nicolas Martin, Associate Professor, Crop Sciences, UIUC

Talk Overview: This presentation delves into a multiyear, on-farm investigation of spatially variable corn yield responses to nitrogen fertilizer. Using high-resolution management zone maps and site-specific soil, topographic, and weather data, researchers applied randomized nitrogen rates across hundreds of subplots at seven locations in Illinois from 2016 to 2021. A random forest algorithm revealed that weather-driven variability was the dominant factor influencing yield responses, overshadowing expected patterns based on soil properties and topography. Temporal inconsistency further highlighted how certain plots that responded positively to higher nitrogen rates in one season did not necessarily do so in subsequent years. These findings underscore the need for more robust temporal data collection and real-time sensing methods to enhance predictive accuracy, refine zone delineation approaches, and ultimately guide adaptive nitrogen management in precision agriculture.

About Dr. Nicolas Martin
Dr. Martin improves long-term profitability and stability of cropping systems by exploring applications of quantitative methods on big data. He leverages interdisciplinary efforts to expand the frontiers of agricultural research; investigates quantitative methods on processes at multiple spatial and temporal scales; and studies effective approaches to implement new insights and discoveries in agricultural decisions and operations.