Research Update: IoT Smart Traps

At a Glance

ETDS Progress

Research Topic: Internet of Things (IoT) Smart Traps and Scouting Aids for Small Insects
PI: Glen Rains
Team: Denis Kiobia (PhD), Canicius Mwitta (Post-Doc), Shekhar Thapa (PhD)
Main Objective: Develop IoT smart trap and scouting aid focused on remote monitoring of small insect pests and beneficials.

Motivation

Current trapping and scouting methods are laborious, time-consuming and error prone
Traps are spatially spread out over large areas and require time and money to manage manually
Small insects are not detected and reported in current IoT traps in the market and are hard to report accurately in the field

Proposed Solution

Develop remote IoT trap for insects underneath leaves
Train machine learning model to identify, classify and count whiteflies, thrips and aphids on trap
Trap data available on a daily basis from website
Handheld camera to ID insects on bottom of leaves
Track population changes across large geographical area

Results to Date

A remote IoT smart trap has been developed using yellow sticky tape, camera, embedded computer and modem. We are currently on version 3 with improvements to the camera, sticky tape alignment and environmental connectors

We are testing a new smart trap design that uses air suction to hold insects instead of sticky tape this season. It also records leaf wetness, air temperature, and humidity output.

A RCNN model was developed to identify, classify, and count whiteflies on yellow sticky tape.

A prototype handheld scouting assistant was tested and initial results will be used for future improvements

Solar powered pest traps

Whitefly identification and counting on sticky tape using RCNN model

Predicted whitefly population from samples whitefly counts on top three nodes of cotton.

Next Steps

Finish development of handheld scouting aid to take images, identify and count insects underneath leaves
Use transfer learning to train RCNN model to detect and count insects on new trap background and on bottom of leaves
Test final trap design at several locations in cotton and vegetable crops

Using handheld scouting aid prototype

Image Taken by Scouting Aid of Underside of Leaf

New Smart Trap Design

Assembly Guide for Small Pest Suction Trap

Citation

Parab, C.U.; Mwitta, C.; Hayes, M.; Schmidt, J.M.; Riley, D.; Fue, K.; Bhandarkar, S.; Rains, G.C. Comparison of Single-Shot and Two-Shot Deep Neural Network Models for Whitefly Detection in IoT Web Application. AgriEngineering 2022, 4, 507-522, https://doi.org/10.3390/agriengineering4020034
Kiobia, D.O.; Mwitta, C.J.; Fue, K.G.; Schmidt, J.M.; Riley, D.G.; Rains, G.C. A Review of Successes and Impeding Challenges of IoT-Based Insect Pest Detection Systems for Estimating Agroecosystem Health and Productivity of Cotton. Sensors 2023, 23, 4127. https://doi.org/10.3390/s23084127

4D Farm