Search for: All records

The concept of City 2.0 or smart city is offering new opportunities for handling waste management practices. The existing studies have started addressing waste management problems in smart cities mainly by focusing on the design of new sensor-based Internet of Things (IoT) technologies, and optimizing the routes for waste collection trucks with the aim of minimizing operational costs, energy consumption and transportation pollution emissions. In this study, the importance of value recovery from trash bins is highlighted. A stochastic optimization model based on chance-constrained programming is developed to optimize the planning of waste collection operations. The objective of the proposed optimization model is to minimize the total transportation cost while maximizing the recovery of value still embedded in waste bins. The value of collected waste is modeled as an uncertain parameter to reflect the uncertain value that can be recovered from each trash bin due to the uncertain condition and quality of waste. The application of the proposed model is shown by using a numerical example. The study opens new venues for incorporating the value recovery aspect into waste collection planning and development of new data acquisition technologies that enable municipalities to monitor the mix of recyclables embedded in individual trash bins. 

Deep neural networks are revolutionizing the way complex systems are designed. Consequently, there is a pressing need for tools and techniques for network analysis and certification. To help in addressing that need, we present Marabou, a framework for verifying deep neural networks. Marabou is an SMT-based tool that can answer queries about a network’s properties by transforming these queries into constraint satisfaction problems. It can accommodate networks with different activation functions and topologies, and it performs high-level reasoning on the network that can curtail the search space and improve performance. It also supports parallel execution to further enhance scalability. Marabou accepts multiple input formats, including protocol buffer files generated by the popular TensorFlow framework for neural networks. We describe the system architecture and main components, evaluate the technique and discuss ongoing work.