With an infinite amount of objects in the world, it is usually hard to keep track and to sort out our waste. In addition, there is still a lot of work that needs to be done in order to protect the planet and to minimize waste.

What it does

Using imagine recognition and machine learning, eTRI will guide you as to sorting your waste. Simply take a picture of an object you wish to sort and eTRI will inform you where to put it: whether it goes in the recycling bin or at a special drop-off point such as Electrobac for electronic devices.

How we built it

Challenges we ran into

  • Standardizing data;
  • Location points in data sometimes not representative of the real location;
  • Tensorflow.js not working with Windows -- only with Linux distros;

Accomplishments that we're proud of

Make an app using image recognition.

What we learned

  • Data Cleaning
  • GeOJsONnnn
  • IBM Watson Image Recognition
  • TensorFlow JS
  • Teamwork

What's next for eTRI

The image recognition could be reused for other application.

Used datasets


  • Collecte du recyclage [HackQC18]
  • Collecte des ordures [HackQC18]


  • Calendrier de collecte des matières résiduelles [HackQC18]

Ville de Québec:

  • Collecte des matieres recyclables [HackQC18]
  • Collecte des dechets [HackQC18]


  • Calendrier de collecte des matières résiduelles par secteur [HackQC18]




  • Secteurs de collecte [HackQC18]
Share this project: