A Distributed Learning and Automation System for Passive Solar Greenhouses

The proposed research focuses on optimizing the partner organization’s innovative passive solar greenhouses. We propose a distributed learning and automation system (DLAS) that collects sensing data from multiple greenhouses. The DLS learns an optimal predictive model by aggregating the local models from multiple edge devices. The DLAS keeps improving the model as more sensing data is available. The predictive model will be used for automating the control of multiple greenhouses. The proposed project will benefit the partner organization by automating and optimizing the greenhouse controls with AI-based predictive decisions at the edge. The proposed project will help achieve energy efficiency and the eventual net-zero carbon emission goal.

Faculty Supervisor:

Steve Drew

Student:

Partner:

FreshPal Ltd.

Discipline:

Computer science

Sector:

Agriculture

University:

University of Calgary

Program: