Skip to main content
ARS Home » Plains Area » Fargo, North Dakota » Edward T. Schafer Agricultural Research Center » Weed and Insect Biology Research » Research » Research Project #444557

Research Project: Fusion of Machine Learning and Electromagnetic Sensors for Real-Time Local Decisions in Agriculture

Location: Weed and Insect Biology Research

Project Number: 3060-21220-032-017-S
Project Type: Non-Assistance Cooperative Agreement

Start Date: Jul 1, 2023
End Date: Oct 31, 2025

Objective:
The overall goal of this multi-year project is to bring together personnel from both USDA and NDSU to develop novel DC-to-Light electromagnetic sensors, unique field-ready platforms to bring sensors to the grower, and fusing them with Machine Learning for real-time decision making. Towards that goal, the project will focus on the following objectives: (1) Partner with USDA and various constituents to conduct fundamental research on new electromagnetic sensor technologies for agriculture; (2) Carry-out an extensive technology readiness level study of existing sensor technologies and apply them to agriculture; (3) Develop Machine Learning (ML) theory specific for edge computing, electromagnetic sensors and applications in agriculture; (4) Implement secure Information Technology (IT) platforms for this initiative; and (5) Align and deploy these new technologies in the field for grower needs.

Approach:
During the course of the second year of the project, the objectives will focus on several crop varieties and insects; namely sugar beets, potatoes, the honeybee, sunflowers and cereal crops (oats, wheat and barley), dry beans and will have the following continued approaches: (1) continue to assess the immediate sensing needs for the aforementioned crop varieties and insects; (2) continue to develop new and survey existing electromagnetic sensors & instrumentation circuitry to conduct the needed measurements, in which the sensors are not limited to imaging, LiDAR and microwave radar; (3) develop lab or greenhouse prototype systems that closely resemble the field-based sensing needs; and (4) further deploy the sensor system in the field for real-time measurements and ML-based decisions.