Collaborative Research: GOALI: AIS gene library based real-time resource allocation on time-sensitive large-scale multi-rate systems

The objective of this project is to investigate and develop a gene library-based resource allocation methodology for effective and efficient time-sensitive large-scale multi-rate system integration over communication networks. The Impaired Driver Electronic Assistance Testbed at the North Carolina Center for Automotive Research serves as an accident prevention demonstration project. The approach is to use the gene library to classify and detect abnormalities in vehicle movements in various traffic environments and to provide optimal real-time sampling rate adaptation and emergency intervention. An artificial immune system is used to optimize the gene library, which can then be used in real time and can adapt to its environment to realized optimal solutions.

With respect to intellectual merit, this research pursues a vertical integration of multiple layers of systems dynamics (from networks, to distributed agents, and to users) with the goal to provide efficient and reliable system operation by considering bandwidth constraints, hybrid structures, and different control topologies for adaptive real-time optimal resource allocation for distributed sensor, actuator, and controller agents over communication networks.

With respect to broader impacts, this project is a synergistic integration of theoretical analyses with industrial applications. Joint research, bidirectional education programs, and commercialization serve as a three-tiered structure for the project. Outreach and dissemination efforts include industrial internships for students, summer high school student experiences, inter-institutional visits, and public seminars. Recruitment of underrepresented groups is actively pursued. The research has the potential to reduce vehicle accident-related costs from more than $250 billion per annum in the United States while enabling older drivers to retain mobility and quality of life.