Overview
Faculty, research staff, and students from Appalachian State University are set to showcase their innovative research at the upcoming IEEE SoutheastCON 2024 Conference. With a focus on advancing technology and addressing critical challenges, their work spans diverse areas, from pioneering bee traffic estimation using advanced machine learning techniques to developing comprehensive data visualization systems for honey bee research.
Publications
Bee Traffic Estimation with YOLO and Optical Flow
Authors: R. Thompson Dickson, R. Mitchell Parry, Christopher Campell, Rahman Tashakkori
Abstract: Honey bees are vital to our food chain as they efficiently pollinate many crops and fruits. With the collapse of an alarming number of honey beehives in recent years, there has been a substantial need for regular monitoring of their health status and population growth. Traditionally, beekeepers have manually examined their hives to assess health and growth rates. Traffic in and out of hives provides significant information about the hive’s health status and population size. The Appalachian Multi-purpose Apiary Informatics System (AppMAIS) project obtains video recordings at the entrance of 28 hives in the Western region of North Carolina, USA and 3 hives in Belgium. This has allowed us to automate the estimation of traffic at the entrance of the hives using Image Processing and Machine Learning tools. This paper provides details on the YOLO-based and Optical Flow-based approaches we have utilized to estimate the traffic in and out of these hives. With an error of only 17 bees in our estimation models, our early results have been promising. Automated traffic monitoring can significantly reduce the number of required manual hive inspections.
 |  |  |  |
BeeVis: A Comprehensive Honey Bee Data Visualization, Exploration, and Analysis System
Authors: Tinghao Feng, Sam Arkle, Christopher Campell, and Rahman Tashakkori
Abstract: Honey bees play an important role in pollinating major crops and fruits across the world. Unfortunately, in recent years, many honey bee colonies have been lost due to various causes that, at some point, were referred to as the Colony Collapse Disorder (CCD). Some research, including the Appalachian Multi-Purpose Apiary Systems (AppMAIS) have acquired a large amount of data such as audio and video recordings as well as humidity, temperature, weight, and genetic data from honey beehives to determine the health of their bees. The amount of data produced by AppMAIS is a good example of Big Data, which is simply impossible to analyze with traditional computing approaches. Such analysis requires sophisticated visualization and exploratory tools to study honey bee behavior successfully and the environmental factors affecting them. This paper will provide details on the design, implementation, and application of a comprehensive visualization system called BeeVis, which has been developed as part of the AppMAIS project for visualizing, exploring, and analyzing honey bee data for 29 hives in the Western Region of North Carolina. BeeVis has assisted us in gaining significant knowledge about the health of the AppMAIS hives. A significant contribution of BeeVis in our research is the ability to provide options for comparison of our hives at different development stages.
 |  | | |
AppMAIS Audio Data Labeling Application
Authors: Christopher Campell, Rahman Tashakkori, Alex Somer, Logan Richardson, and Aedan Simons-Rudolph
Abstract: The Appalachian Multi-purpose Apiary Informatics System (AppMAIS) research project has monitored the behavior of some honey beehives in a small region of Western North Carolina in the past two years. With over two million audio files sampled across twenty-nine research hives, manually labeling each recording was infeasible. The presence or absence of piping, a bioacoustic signal emitted by honey bees as a precursor to a swarm event provides critical information for beekeepers. The AppMAIS Audio Data Labeling Application (ADLA) was developed in our lab to label the vast audio data. To date, over thirty-eight thousand recordings have been labeled using this application. We were able to train a Machine Learning model to detect piping with high accuracy. This paper describes the implementation of the labeling application and its user interface in detail. The application has helped us identify several swarms and backtrack the chain of events several days before they occurred, allowing us to utilize our trained model as a preventive measure.
| |  |  |  |
CausalBO: A Python Package for Causal Bayesian Optimization
Authors: Jeremy Roberts and Mohammad Ali Javidian
Abstract: This paper introduces CausalBO, a Python package developed to enhance the applicability and utility of the Causal Bayesian Optimization (CBO) algorithm. The original CBO algorithm, developed by Virginia Aglietti et al. [1], integrated causality into Bayesian optimization to address its limitations in complex, interconnected systems like healthcare. However, the initial implementation had several drawbacks, such as specificity to certain datasets, non-modularity, and complex setup requirements. CausalBO addresses these issues by offering increased readability and usability, integration with widely used libraries like BoTorch and DoWhy, and simplification of the CBO loop. It is designed to be user-friendly, catering to those with limited knowledge of causality and do-calculus. The effectiveness of CausalBO is demonstrated through case studies, including a synthetic experiment and a healthcare scenario, showcasing its applicability and versatility.
 |  |