Smart object detection based on an advanced network of cameras and a Machine Learning algorithm.

Camera, edge-computing, op-maat constructie, PWA, ReactJS, Microsoft Azure


At Recupel, conservation of the environment is the most important factor, so they work together with recognised processing companies. These processing companies were selected on the basis of their experience with ecological processing, their efficiency and their cost price. Recupel is also responsible for depollution. Depollution stands for the separation and removal of the most harmful substances. In this way, this organisation contributes to a more beautiful world.

Recupel collaborates with a social employment centre for the classification of all recycled electro. This process is done manually, to make it run more smoothly, they were looking for a new efficient way.


We built a full setup with smart cameras, a scale and software (Microsoft Azure)

How does it work?

  1. The electro is placed on the scales and the cameras draw a picture from each view.
  2. The captured photos go through the artificial intelligence software on the Azure Cloud.
  3. The object will be placed with weight and photos under the correct classification.

In case of faulty recognition, for example between a laptop and mobile phone charger, you can adjust the result manually. This makes the algorithm smarter through machine learning and it will recognize it effortlessly in the next similar object.

We were looking for an efficient innovative solution. By involving Artificial Intelligence, the whole process is much smoother and very accurate.

Dare to challenge us?

Discover some other projects