Back to Home
Techwaste

A Smart Solution for Managing Electronic Waste with AI

9/10/20231 views
Techwaste Application Overview

Problem

In 2021, Indonesia generated an estimated 2 million tonnes of e-waste, the largest volume in Southeast Asia. Frequent replacement of electronic devices and rapid technology advancement exacerbate waste accumulation.

Many people remain unaware of how and where to dispose of their electronic waste properly. Existing recycling solutions are fragmented, often inaccessible, or lack education and technical integration.

Solution

Techwaste proposes a 3R (Reduce, Reuse, Recycle)-focused approach using machine learning and cloud computing. It enables users to:

  • Identify types of electronic waste using AI-powered image classification.
  • Access educational content on proper disposal and recycling methods.
  • Connect with recycling experts and local collection points through a forum marketplace.

Highlights:

  • AI-based E-waste Detector: Uses TensorFlow and Hugging Face to classify up to 15 categories of electronic waste.
  • Cloud-Native Infrastructure: Deployed with Google Cloud Platform (GCP) for scalable compute and storage.
  • Community Integration: Offers article sections, discussion forums, and expert-led insights for sustainable e-waste management.
  • High Accuracy: Achieves 98% classification accuracy, producing predictions in under one second.

tldr

  • Situation: Indonesia faces the region’s highest electronic waste volume with limited awareness of proper recycling methods.
  • Task: Create an AI-driven app to identify, educate, and connect users with e-waste recycling solutions.
  • Action: Built and deployed a TensorFlow-based classifier with cloud integration via GCP.
  • Result: Delivered a scalable, high-accuracy app that promotes recycling and sustainability through accessible technology.

Technical Highlights

  • Machine Learning: TensorFlow for model training and evaluation.
  • Cloud Infrastructure: Google Cloud Platform (Compute Engine, Cloud Run, CloudSQL, Cloud Storage).
  • Model Hosting: Hugging Face for API deployment and inference.
  • Backend Logic: Python-based ML pipeline integrated with REST APIs.
  • Team Collaboration: Developed as part of Bangkit 2023 cohort team (C23-PS324).
Techwaste Mobile App Interface
Techwaste Cloud Architecture Diagram
Techwaste Promotional Poster
Techwaste Promotional Poster
Techwaste Logo

Reflection

Techwaste combines AI, sustainability, and education — proving how machine learning can accelerate environmental awareness. The project experience reinforced the significance of cross-disciplinary teamwork, involving AI engineers, UI/UX designers, and business strategists to address real-world ecological issues.

Future improvements:

  • Integration with Google Maps API to display nearby recycling drop-off locations.
  • Expansion to support plastic and paper waste classification.
  • Launching a public API for NGOs and government sustainability programs.