Decentralized Autonomous Electric Mobility-on-Demand Services for Individuals with Physical and Cognitive Disabilities
MetadataShow full item record
This report discusses the foundation of innovative decentralized mobility services for individuals with physical or cognitive disabilities using disability-friendl,y autonomous electric mobility on-demand (AEMoD) services. By leveraging both the Internet-of-things (IoT) and its associated fog control capabilities, this framework will enable real-time, localized, autonomous, disability-aware, and battery-level-based dispatching and charging decisions for a fleet of AEMoD services distributed in multiple city zones. Through IoT-enabled fog control, computational resources are pushed closer to customers in each city zone, which enables the collection of real-time information about the AEMoD services, tracking of their state-of-charge, and the collection of service requests from customers. Driven by these collected data, these fog controllers will provide decentralized, efficient, and responsive dispatching and charging decisions within each zone to serve customers, prioritizing their needs in a timely manner, while maintaining suitable AEMoD state-of-charge for subsequent trips. The proposed fog-based architectures for localized AEMoD system operations provides a good solution for the communication/ computation delays that restrict massive AEMoD operations in big cities. These emerging architectures will soon become widely used, allowing all localized operational decisions to be made with very low latency by fog controllers located close to the end applications (e.g., each city zone for AEMoD systems). The proposed architecture also employs an optimized, multi-class charging and dispatching queuing model, with a partial charging option for AEMoD vehicles, to provide the best solution to the AEMoD charging delay challenges for each zone. The stability conditions of this model and the optimal number of classes were derived. The decisions on the proportions of each class of vehicles to partially/fully charge or directly serve customers are optimized to minimize the maximum and average system response times by using convex optimization and Lagrangian analysis. Analysis results showed the merits of our proposed model and optimized decision scheme in comparison to both the always-charge and the equal-split schemes. Furthermore, the comparison between the maximum and average problem solutions exhibited negligible variance, which favored the use of the maximum solution because of its lower complexity.