For resource limited cities, garbage collection is a challenge. People with less social interest or without an alternative, may simply dump their garbage at public places. Even if designated garbage disposal sites are available, the city authorities may need to prioritize garbage collection. For this prioritization, a monitoring and management solution is required. Images below are examples of how waste is present at public places and how city authorities perform collection.
A computer vision based monitoring system can be the first step. A system in which a camera, mounted on top of a bus (or any other vehicle), can capture images with the location information. These images can then be analyzed by machine learning algorithms to find the amount of garbage present in each image. Thus, by using such an algorithm, the places where garbage is present can be found. What would also be interesting is to estimate the amount of garbage from the images.
The information of how much waste is present at which city locations, can be used for management. A tool, e.g. a scheduler can be used to create routes for the workforce to collect the garbage. Automation of the scheduling together with monitoring will allow city authorities to dynamically allocate resources.
Problem owner: Waste management authority of Lahore.
Problem domain: Internet of things, monitoring, programming, machine learning.
Is funding possible? Yes, funding is available for the technical realisation of the project (e.g. hardware costs).
Duration: At Least 6 Months.