CAREER: Enabling Adaptable, Object Oriented, Real-time Systems

The increasing need for real-time guarantees in mobile embedded systems poses a unique set of challenges, including changing software, just-in-time (JIT) compilation, and dynamic interactions between applications of differing criticality levels. This project develops techniques to provide adaptability to software systems that require stringent correctness and timing guarantees. The novelty of this work is in the development of new techniques to ensure the correctness and timeliness of adaptive software, new programming models to ease the development burden of adaptive software, and the development of an execution platform for adaptive software. The impacts of this work will be felt by providing predictable execution in mobile applications for sensing, personalized health care, command and control, and automotive domains, as well as providing professional development on the topic of mobile programming to teachers in the Buffalo Public School district. Additionally, this award supports the training of Ph.D. students and integration of research into educational materials. This project investigates language level, compiler, and run-time techniques that enable the construction of adaptive real-time systems. The project develops a novel, partitioned run-time, including support for executing multiple real-time and non real-time applications, cross application communication abstractions with concrete predictability guarantees, as well as partitioned virtual machine-level services, such as JIT compilation and garbage collection. The run-time is expressive enough to support Android's APIs and execution model, yet affords predictability guarantees necessary for real-time support. Furthermore, the project develops a series of validation mechanisms that ensure real-time guarantees for mobile systems, and constructs a programming model that expresses adaptability in real-time systems. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.