The problem Rhinestone solves

Our project aims to automate the process of urban waste segregation and facilitate efficient management using digital image processing, cascade classifiers, electromechanical systems, all linked, controlled and coordinated by a local server. The project is a concept demonstration with both hardware and software. First stage of the project deals with automatic segregation of waste materials into metallic, biodegradable and non-biodegradable categories, this is done by a single board computer that runs an image processing algorithm with trained models accordingly. The same single board computer is acting as a high speed local server that facilitates data transfer between an on-board sensor hub and s smartphone which can be used to monitor several parameters that include the presence of toxic materials, over-weight and over-flow in the waste containers and so on. The main problem that this project addresses is the sheer mismanagement of waste in metropolitan cities and inefficient segregation due to which waste recycling plants and smelters have to run for extra time durations, this causes immense wastage of engergy and resources. Our project brings the speed and accuracy of portable computing for enhancement of waste handling methods and conservation of energy.

Challenges we ran into

Following are the challenges that we ran into while building this project:

1. Connecting multiple development boards with a local server.
2. Interfacing different microcontrollers and microprocessor boards.
3. Tuning computer vision pattern recognition parameters.
4. Power management and power distribution.
5. Thermal throttling in Raspberry Pi.
6. Smartphone application integration with local server.