A Simulation-as-a-Service (SIMaaS) Cloud Computing Platform for Traffic Management

Abstract

Traffic simulation is defined as a tool to replicate the real-world condition, produce the possible scenario, and facilitate scientific decision making, which can be divided as two categories: planning and operational management. However, the unavailability of vendor-supplied software, difficulty of finding sufficient staff, a lack of community awareness are the major factors impede the adoption of advanced techniques in simulation. This dissertation developed a Simulation-as-aService platform (SimaaS), based on cloud computing technology, which can seamlessly integrate transportation simulation engines with flexible cloud-based functional modules like network editor, modeling compiler, cloud middleware, and metrics dashboard to reduce the technical burden of staff in traffic management. A mixed integer linear programming-based task scheduler is embedded in the cloud middleware to improve availability of simulation engines. To facilitate the social involvement, the platform breaks down the service life cycle into three stages for the activity modeling of participants and proposed two business models to underpin activities of various stakeholders. Two sample service cases demonstrate the capability and extensibility of the proof-of-concept. The contributions of this study include: II • Architect a SimaaS framework inherent with a business potential to bridge the gap between transportation simulation providers and end-users (i.e., transportation agencies and consulting firms) at different project stages; • Develop a fully scalable and flexible cloud-based network editor and compiler to transform various customer needs into network configuration compatible with simulation engines of various types; • Develop a back-end mechanisms and algorithms for real-time multiple simulation engine deployment and scheduling; • Summarize scalable technical architects for interactive dashboards for simulation results aggregation, interpretation, visualization, and decision support; • Identify, design, and implement three critical stages in service platform modelling for stakeholders. Propose business models for service providers and the platform for developing an evolutionary crowdsourced knowledge-base module to benefit the research and professional community continuously; and • Implement and test the prototype SimaaS cloud-based platform for real-world applications.