Optimizing Operation of Automated Waste Collection Systems in Smart Communities

Automated vacuum waste collection (AVWC) systems employ underground pipe networks and air pressure to transport waste to a central facility. To reduce energy consumption for AVWC systems, we need to minimize the number of waste collections. However, reduced collections will result in high waste accumulations that negatively affect residents' life quality. In this paper, we investigate the impact of different collection schedules on the AVWC system and propose an optimization solution that chooses the collection schedule which minimizes energy consumption, wear-and-tear, and reduces the retention of waste in the system. A straightforward way to model the system would be to model each waste inlet as a decision variable but doing so results in an explosion of the solution space. Our framework overcomes this limitation by grouping waste inlets together based on location and waste arrival profiles. We obtain operational logs from a real-world AVWC system and feed the information gleaned from the logs into an AVWC simulator that we developed. We apply our optimization solution on the system and use our simulator to evaluate the performance of our optimized schedule. We compare our solution against various benchmarks and find that our solution saves 14.4 % energy consumption and reduces waste retention by 6.9 %.