Did you know Malta generated almost 3 million tonnes of solid waste in just one year? That’s nearly 6 tonnes for every resident. A lot of it ends up in places we can’t even see — cliff sides, nature reserves, remote beaches…
That’s why a new interdisciplinary AI-powered project from the University of Malta is taking waste detection to new heights — literally.
Funded by the Xjenza Malta's Technology Development Programme-Lite (TDP-Lite), AWIGS (Aerial Waste Identification and Geolocation System) is pushing the boundaries of how technology can support environmental preservation. In collaboration with the CMD (Cleansing and Maintenance Division) department, a team of researchers is exploring the use of drones equipped with computer vision to detect litter in areas that are otherwise inaccessible or difficult to monitor, such as nature reserves, cliff sides, remote beaches, and Natura 2000 sites.
The core advantage of using drones lies in their ability to cover large areas in a single flight, drastically reducing the need for manual land traversal. This approach not only increases efficiency but also helps safeguard ecologically sensitive environments like garigues, where foot traffic could cause more harm than good. Additionally, drones can safely inspect hazardous or precarious locations where it would be dangerous for people to venture on foot.
At the heart of the system is a drone operated either manually or via a predetermined flight path, flying at a maximum altitude of 30 meters above ground level. The drone captures high-resolution 4K images, which are then transferred to CMD’s control room for analysis. Initially, data will be physically delivered via SD cards, but the team is developing an Android app to automate uploads using Wi-Fi or mobile data (5G).
Once received, the imagery is processed by custom-built object detection algorithms, developed using AI computer vision techniques. These models are being trained specifically on visual data from Maltese landscapes, ensuring that the system can recognize all forms of litter found locally — from construction debris and bulky refuse to common everyday waste.
When litter is detected, it is marked on a digital map (similar to Google Maps) using the geo-coordinates embedded in the drone images. This allows CMD administrators to visualise the spread of waste in real time and dispatch cleanup teams efficiently with the appropriate tools and vehicles.
The project is being led by Dr Dylan Seychell, with key contributions from Gabriel Hili and Matthias Bartolo (Research Support Officers), as well as academics Dr Konstantinos Makantasis, Prof. Matthew Montebello (Head of the Department of Artificial Intelligence), and Prof. Ing. Carl J. Debono (Dean of the Faculty of ICT).
Looking to the future, Dr Seychell envisions expanding this line of research even further. His team is exploring the potential of mounting cameras on CMD’s sweeper robots to detect improperly placed garbage bags, and is also investigating robotic arms for drones — with the long-term goal of enabling drones to not just identify, but also collect litter autonomously. While these ideas remain in the conceptual phase and are currently unfunded, they represent an exciting frontier in the integration of AI, robotics, and environmental stewardship.
This project exemplifies how technological innovation can play a transformative role in addressing environmental challenges, and signals a promising future for automated waste detection and collection across Malta and beyond.
