Utilizing Remote Sensing and Machine Learning to Improve Poultry Farm Productivity

Within the Canadian context, poultry farmers are constrained by regulations that predetermine chicken prices and market supply. As a result, they are limited in the approaches they can take to improve the profitability of their operations. Within this regulatory framework, farmers must rely on measures that can be applied on their farms to improve chicken’s growth performance while reducing production costs. In this project, we aim to find the effective approach to utilizing remote sensing and machine learning to improve poultry farm productivity. Specifically, we would like to identify which network/sensor configuration is best suited to meet remote sensing needs on rurally located chicken farms. In addition, we plan to design a machine learning based scheme to analyze data from farm sensors in order to identify any environmental concerns (e.g. barn humidity and/or temperature is too high or low) and to suggest actions for poultry farmers accordingly. Finally, we attempt to assess how data collected from remote sensing equipment can be used to increase performance and/or growth of chickens.

Faculty Supervisor:

Qiang Ye;Deborah Adewole

Student:

Yitong Zhou;Taiwo Makinde

Partner:

myFlock

Discipline:

Animal science

Sector:

Agriculture

University:

Dalhousie University