Student Spotlight on Tyler Seutter: Why not What: Using AI to improve Tree Fruit Pest Management

Pennsylvania is one of the top apple and peach-producing states. The annual value of these crops reaches $130 million for apples and $20 million for peaches, grown across approximately 25,000 acres. Pests like the Oriental Fruit Moth and Codling Moth are major threats to these valuable fruit crops. Oriental Fruit Moths burrow into fruit shoots, causing dieback of new growth. Unfortunately, this is often detected too late, resulting in significant crop losses and substantial economic impact. Traditional monitoring methods rely heavily on labor-intensive manual inspections, which can be inefficient and prone to errors. Therefore, the need for more efficient and accurate monitoring methods is increasingly evident.

Artificial intelligence (AI) tools can revolutionize pest management by enabling real-time automated detection and precise identification of pest populations through the use of automated traps and machine learning (ML) algorithms. AI has the potential to predict pest outbreaks, optimize pesticide usage, and implement timely interventions, reducing the need for costly, labor-intensive manual inspections and increasing safety for growers.

Tyler Seutter, a PhD student in the Department of Entomology at Penn State University, aims to discover the underlying developmental differences in Oriental Fruit Moth and Codling Moth populations across the tree fruit growing regions of Pennsylvania. Tyler’s work utilizes machine learning to develop predictive pest development models, providing growers with more efficient and cost effective tools for their pest management plans.

“My goal is to use the data that we obtain from traps to model how different weather and landscape features influence moth populations so we can help growers predict when pests arrive in their farms,” Seutter explains.

Seutter manages around 250 smart traps that are focused on AI monitoring. These traps are built by CropVue, and FMC provides an app that gives users easily digestible data in the form of capture numbers, tables, and charts. (CropVue is a precision agriculture company, and FMC Corporation is an agricultural science company.) While conventional traps are standard sticky traps, new, automated traps are making their way into the field. These traps often feature remote monitoring and data transmission, automated species identification using AI, and environmental sensing capabilities.

Currently, growers rely on pest management models that are over 40 years old, with minimal efforts in tracking regional variations in pest populations. Seutter faces the challenge of confronting these outdated models and reevaluating them with the latest technology. By using pheromone traps equipped with AI-powered monitoring tools, Seutter is collecting precise moth capture data, enabling a more detailed comparison and highlighting the need for region-specific pest management strategies.

Addressing these challenging problems with technology requires a human perspective. “I need to go after the why, not just the what,” Seutter explains. Seutter hopes to identify if there are regional variations and if so, pinpoint the exact contributing factors. With this information, Seutter can develop tailored models specific to particular regions on Oriental Fruit Moth and Codling Moth development timing, efficient pesticide spraying behaviors, and adult moth emergence. Seutter aims to make pest management easier and more convenient for growers, ensuring timely and cost effective pest control. “I want to help real people, the ones growing our food, and I know my research will make a tangible difference in their lives,” Seutter explains.

Tyler Seutter is a fellow in the INSECT NET Training Program. Tyler Seutter is a PhD student mentored by Professor Grzegorz Krawczyk in the Department of Entomology. Seutter’s research is also funded by a fellowship awarded by FMC.

This article was written by Micah Delattre, a fellow in the INSECT NET program and a MS student mentored by Bo Cheng in the Department of Mechanical Engineering. It was prepared as part of the INSECT NET Science Communication workshop series coordinated by Drs. Christina Grozinger and Natalie Boyle during summer 2024.