Recipient Organization
VECTECH, LLC
1812 ASHLAND AVE STE 100
BALTIMORE,MD 212051546
Performing Department
(N/A)
Non Technical Summary
Native bees are a critical part of our nation's ecological infrastructure. While much of media coverage is centered on support for honey bees, which are imported and domesticated, native bees are more effective pollinators than honey bees for many plants, while also being more vulnerable due to their reliance on dwindling native ecosystems. Despite native bee's superior effectiveness, most agriculture practices rely on imported honey bees for pollination, often creating more pressure on native bee populations. Part of the reason for this is a knowledge gap in the existing pollinating services in a region; which native bees are in my area, which might support pollination of my crops, and how do I attract them to my fields? While some researchers attempt to fill this knowledge gap through trapping, crowdsourcing, and direct observation, these methods have not provided sufficient data for most species and most regions. This leaves farmers that have sustainability goals incapable of targeting the right species for their region and crop(s) of choice.To this end, this project aims to develop a smart native bee camera trap to provide this critical information in an accessible way, supporting both sustainable agriculture and conservation efforts. First, we will define the necessary quality of the imaging system for bee species identification, then develop the imaging system itself. We will make a modular attachment to the trap, which will attract native bees to land on a specific spot by imitating a flower. When bees approach this attractant, the high resolution video will begin. This will result in a preliminary system fit for use by native bee researchers and other entomologists. Development of the artificial intelligence methods for native bee species identification, which will make the system accessible to a lay person, is intended for the phase II portion of this work.
Animal Health Component
15%
Research Effort Categories
Basic
15%
Applied
15%
Developmental
70%
Goals / Objectives
The goal of this project is to develop a high-resolution camera trap for imaging and species identification of native bees that enhances capabilities in environmental monitoring to support both conservation and sustainable agriculture practices. Objective 1) Design and validate an optical configuration for high-resolution imaging and species identification of pollinators. Objective 2) Tailor a pollinator attractant for compatibility with the optics system, and modular attachment method for use of alternative attractants. Objective 3) Develop a methodology for detection of an insect on the imaging stage and pollinator attractant.
Project Methods
The nature of the effort in this proposal is solely technical development and evaluation of the outputs of that development. The resolution requirement for species identification of native bees will be focused first on native bees in Maryland. The optical requirements will be defined by physical measurement of key identifying features of target bee species using online databases or physical specimens. This will result in a resolution requirement defined by resolvable microns and depth of field requirements, which will be used to develop a remote monitoring system achieving those specifications, and capable of field operations. The imaging system will be evaluated through early data gathering in various outdoor environments typical of expected use. After confirmation of optics system viability in varied lighting conditions, a modular attachment for different visual pollinator attractants will be developed, to support future management of data biases in native bee monitoring. Further outdoor data gathering will evaluate the performance of rudimentary attractants in bringing native bees to the attractant and imaging stage. The resulting video data from the optical system during these experiments will be used as data for developing a computer vision methodology for activation of the high resolution video of the imaging stage to support identification. Evaluation of this methodology is slated for development in a potential Phase II of this work.