Trash-Sorting Concepts

We know the problem

The 'Auto-Trash' concept is a way to make waste disposal more efficient, which should be a priority for anyone interested in slowing the effects of climate change. Designed as a proof of concept for the TechCrunch Disrupt Hackathon, Auto-Trash is able to identify whether an object is recyclable or compostable and then place that item in the correct bin.

To make its assessment, the Auto-Trash machine uses a Raspeberry Pi computer that takes an image of the object placed on the can's rotating lid. An algorithm powered by Google's TensorFlow AI engine then judges what type of item the image is based on its machine learning tenets. If the item is recyclable, the lid tilts left and drops it in the recycling, and vice versa for the compost.