StreamSight: General Purpose Algorithm for Estimating Waste Weight at the Waste Collection Stage

The problem at hand is an inability for municipalities across Canada to audit recycling and waste collection. Auditing comprises both the quantity and type of objects being disposed or recycled. The only existing option is manual, which cannot be applied beyond small intermittent inspections due to the significant cost, time commitment and inefficiency.
In this project, two research directives are targeted: (1) to classify items at the point of waste collection in real time, and (2) to estimate the weight of individual collection bins using information obtained in (1), the fuel consumption in vehicles, and motion data obtained while loading waste into the vehicle using machine learning techniques

Faculty Supervisor:

Mehran Mehrandezh

Student:

Niloufar Malekpour

Partner:

Prairie Robotics Inc.

Discipline:

Engineering

Sector:

Administrative and support, waste management and remediation services

University:

University of Regina