Hoooly – AI & Robotics for waste monitoring and sorting

Ganiga turns waste into data. Today, over 60% of municipal waste is still disposed of without sorting; trash bins haven’t evolved in 50 years, and there’s no oversight of contamination—this makes it difficult to comply with ESG, CSRD, and PPWR regulations and drives up disposal costs.

Hoooly is Ganiga’s AI infrastructure for smart waste management, based on a proprietary computer vision algorithm capable of recognizing any object or food item from a single image with 97% accuracy in less than a second, measuring weight, volume, material, and brand. The data is transmitted in real time to a dashboard that generates certifiable KPIs (sorting rate, CO₂ avoided, volumes by category), ready for ESG audits and reductions in TARI and the Plastic Tax.

We present three solutions from the Hoooly family:

Hoooly Food is an AI micro-camera that can be installed in existing trash bins in food service areas, designed to monitor food waste, calculate its economic impact, and provide data to optimize meal preparation and kitchen management.

Hoooly Small is a smart trash can with three doors that recognizes the type of waste placed near it and automatically opens the correct compartment (paper, plastic, general waste). It is equipped with a display that transforms it into a digital touchpoint: it shows real-time data and insights to users and, during periods of inactivity, can be used as advertising space or for corporate communications.

Hoooly Indoor is a smart trash can with a single door, designed for high-traffic environments (airports, train stations, offices, shopping malls, hotels, food service areas, entertainment venues): it automatically recognizes the material and sorts it on its own, providing visual feedback to the user via the display and notifying operators of the fill level.

The measurable impact of the Hoooly family includes up to a 75% reduction in unsorted waste, a 50% reduction in CO₂ emissions, a 50% reduction in operating costs, and less than 5% foreign material, thanks to a proprietary dataset of over 12 million images and more than 160 devices already installed in the field.