Waste collection has become a major issue all over the world, especially when it is offered as a service for businesses; unlike public waste collection where the parameters are relatively homogeneous. This industry can greatly benefit from new sensing technologies and advances in artificial intelligence that have been achieved over the last few years. However, in most situations waste management systems are based on obsolete technologies, with a low level of interoperability and thus offering static processes. The most advanced solutions are generally limited to statistical, non-predictive approaches and have a limited view of reality, making them weakly effective in dealing with day-to-day business issues (overflowing containers, poor quality of service, etc.). This paper presents a case study currently being developed in Luxembourg with a company offering a business waste collection service, which has a significant amount of constraints to consider. Our main objective is to investigate the use of multiple waste data sources to derive useful indicators for improving collection processes. We start with company-owned historical data and then investigate GPS information from tracking devices positioned on collection trucks. Furthermore, we analyze data collected from ultrasonic sensors deployed on almost 50 different containers to measure fill levels. We describe the deployment steps and show that this approach, combined with anomaly detection and prediction techniques, has the potential to change the way this business operates. We also discuss the interest of the different datasets presented and multi-objective optimization issues. To the best of our knowledge, this article is the first major work dedicated to the world of professional waste collection.
This is an open access article distributed under the Creative Commons Attribution License
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited