High-Level Project Summary
We create a model of the recyclable materials being made and sold as well as match how much is being collected at various trash cans and recycling bins as well as other disposal locations. The remaining uncounted amounts are most likely to find themselves in nature as well as wind up in waterways after a certain amount of time if the location is near tributaries and other connected water systems. Through the analysis of the various sections of the larger system and sensing at collection points we are able to guide in the understanding of how the bigger picture around plastics is forming.
Link to Project "Demo"
Link to Final Project
Detailed Project Description
We have a selection of material column compositions for the scaled trashcan.
We tested a can (represented by a plastic cup because most trashcans around the city were also made of some form of plastic) as well as filled it with paper, plastic, and aluminum (in foil form). We also tested a combination of the plastic bag and aluminum to see how the combination would compare with the separate materials.
Through the understanding of how four kinds of standard waves (Sine, Square, Sawtooth, and Triangle with frequencies played logarithmically from 1 Hz to 20 kHz) pass through the material columns.
We then used Teachable Machine to make a simple model which could be shared online with Tensorflow as the machine learning framework.
Based on how the sound reverberates through the columns and eventually through the collection points we will be able to see the composition of the materials in the containers.
Materials have various energy absorption properties with the bulk material being representative of the components.
In addition to sensing at individual collection points we would collect sales data from local businesses and industry plants to understand how much plastic is being brought into a system for a more robust model of the contributing sources of plastics reaching the environment and oceans.
Using Open Data portals we will then be able to show the amounts of various kinds of plastics entering as well as leaving a community given the sensor data and the industry sales data.
Space Agency Data
None currently.
Hackathon Journey
We hack in many ways though this is our first hackathon as a group. Finding datasets and modeling tools for the industry data was a hurdle.
References
Teachable Machine: https://teachablemachine.withgoogle.com/train/audio
Global Judging
This project has been submitted for consideration during the Judging process.