Lack of awareness of waste segregation and management among common folks. Which is the one affecting factor which contributes towards global warming. There are many categories of waste such as:
- Dry waste
- Wet Waste
- BioMedical Waste
- Nuclear Waste and etc
Unfortunately these waste does get mixed at the source and finally end up at the landfill which makes it very hard to re-segregate, filter or even properly disposal of the waste.
Did you know?: Plastic should never end up in the landfill but it has to be recycled and reuse to reduce carbon footprint in manufacturing other plastics or can be recycled into fuel.
What it does
Aim to tackle this problem the major solution projected on segregation at source and general awareness of use of bin. My modal is trained with little to fewer dry (plastics and similar dry waste) and wet( foods & fruits) which on encounter suggest which bin to discard or dispose. Reducing mixing of waste which can effectively recovery, reusable and recyclable.
How I built it
I trained the modal in tensorflow and google colab. And exported the modal into tensorflow.js modal and hosted at azure node platform.
Challenges I ran into
Many challenges in training and importing the modal and still have to try this modal in the mobile platform and embedded device to make smart-bins.
Accomplishments that I'm proud of
Was able to train the modal and able to label dry and wet waste.
What's next for Whichbin
RoadMap for WhichBin:
- Regress training for other type of waste according to the international color code.
- Building a smartbin with embedded modal.
- Looking for grants and CSR funds for the development.