What does the Distribution-and-Collection-Planning as-a-service (CFG) do?
The KVA Thurgau association and its partners operate approximately 1400 underground containers (UFC) for regional waste collection. However, the previous rigid route concept of waste collection leads to inefficient trips, with some UFCs being half empty and others overfilled. This causes unnecessary costs and burdens the environment. The approach of equipping the UFCs with sensors that display and transmit the fill level has, however, shown in tests that these sensors are comparatively prone to errors, expensive, and also not capable of making predictions: they only show the current fill level of a UFC. In contrast, standard weighing with each emptying forms the basis for creating precise fill level forecasts.
The DCP software (Distribution-and-Collection-Planning as-a-service) from Thurgauer CFG AG enables dynamic waste collection by calculating the daily emptying requirements of underground containers and optimized route planning taking into account logistical restrictions. This innovative approach enables the transition to a more flexible and efficient waste collection, not only for underground containers but also for other containers.
The 2 crucial functions in the data-generating process
The data-generating process forms the basis for precise and efficient delivery and collection planning. It is based on the evaluation of historical data and the use of state-of-the-art statistical methods (artificial intelligence) to ensure optimal resource planning and utilization.
Cloud-based delivery and collection planning as-a-service with s-peers AG
s-peers AG, exclusive distribution partner of the solution and responsible for operations in the Google Cloud.
The cloud-based software service developed by Thurgauer CFG AG enables demand-based, dynamic and tour-optimized waste collection. Thanks to this technology, locations are only visited when the container capacity is exhausted.
The solution is based on fill level forecasts, which are created using an AI approach. These forecasts are based on historical weighing or sensor data and calculate the optimal emptying times for each collection point. The collection points are geographically grouped into clusters to optimize daily route planning. Various logistical constraints such as priorities of the collection vehicles, availability of the vehicles, operating times of the drivers, emptying times and calendars are taken into account.
DCP was developed in R to efficiently apply statistical methods. The concepts and methods are based on mathematical statistics and operations research procedures.
The planned routes are directly transferred to the vehicles' navigation systems. Specifically, this service enables the innovative step from today's static waste collection to the dynamic and flexible collection of the future. Through the use of state-of-the-art technologies, CFG AG supports municipalities and companies in making their waste management smarter and more environmentally friendly.
DCP is available as a service and is seamlessly integrated into the process design of users. This service is provided via the Google Cloud Platform (GCP). A separate container is available for each project and user, housing both the software and the data.
Source code management and release maintenance are done via GitHub.
Ready for the next level? Learn more about the Google Cloud Platform.
Download the free white paper "Distribution & Collection Planning as-a-Service" now
The 4 steps of refuse collection planning
The 4 main added values of this solution: From static to dynamic refuse collection
Learn more on the CFG page now!
