Smart waste
Smart waste collection with integrated sensors. Forecasting fill levels and optimizing collection routes using real-time data.
COMPANY
NORS Group S.A.
RESEARCH CENTRE
LEMA and Universidade do Porto (Faculdade de Ciências)
PRODUCTIVE SECTOR
Logistics and Transport
Problem description
Sotkon, part of the Nors’ Ventures segment and specialized in technological solutions for waste management, installs sensors in recycling containers to monitor fill levels in real time. Using this data, the challenge was to plan truck routes that improve waste collection efficiency, by ensuring waste is collected exactly when and where it is needed.
Challenges and goals
- Forecast container fill levels and anticipate daily collection needs.
- Manage daily fluctuations in container fill levels.
- Plan efficient routes, respecting logistical constraints.
- Minimize costs and environmental impact.
Mathematical and computational methods
The project started by developing a mathematical model to predict the filling levels of recycling containers in the forthcoming hours. The model was used to plan waste collection more effectively through the TOPSIS method, which prioritizes collection based on factors like urgency and location.
The second stage of the project relied on an algorithm that runs three different routing methods — Clarke and Wright, Genetic Algorithm and Ant Colony — to return the best solution. Since the collection points change each day, this approach uses the strengths of each method to find the most efficient solution every time.
Results and Benefits
The collaboration yielded two key benefits: the anticipation of collection needs and the optimization of resources. By prioritizing the most suitable containers, efficiency increased, and the routing algorithm reduced the distance travelled by about 11.5% per route.
Additionally, the percentage of containers collected that were filled above 80% — the company’s ideal threshold — increased from 21.8% to 51.3%, ensuring that the collection is made primarily where it is most needed.
Overall, the project reduced costs, minimized environmental impact, and improved service quality.
