GGrantIndex
← Search

CNS Core: Small: Boomerang: A Symbiotic Software Architecture for Real-Time Distributed Embedded Systems

$497,693FY2020CSENSF

Trustees Of Boston University, Boston

Investigators

Abstract

This project focuses on the development of a real-time distributed embedded system (RTDES) for timing and safety-critical applications, in areas such as semi-autonomous and driverless cars, robotics and manufacturing. The proposed system, called Boomerang, aims to enforce spatial and temporal isolation between tasks of different criticality levels on the same hardware. This will ensure tasks with a severe consequence of failure are isolated from less important tasks, while being able to meet their timing requirements. Boomerang uses a collection of lightweight partitioning hypervisors running on each host to securely divide machine resources among different guest domains. It will support the co-existence of a real-time operating system (RTOS) and one or more legacy systems such as Linux or Android on each physical host, to provide a mutually beneficial symbiotic relationship for the task at hand. The legacy systems will gain access to real-time capabilities, while the RTOS will be empowered with a rich collection of features that would take years to develop. Boomerang is being evaluated in the context of DriveOS, an automotive system for a high-performance, potentially autonomous, electric car being built with an industrial partner. Beyond DriveOS, Boomerang will provide opportunities for new classes of cyber-physical applications in smart manufacturing, robotics, and transportation systems, among others, where decisions are made by machines without recourse to human inputs. Boomerang will offer opportunities to build new distributed systems for neuromorphic computing applications, allowing real-time task pipelines to process large-scale sensor inputs and produce outputs similar to how a brain communicates between sensory and motor neurons. Lessons learned will lead to more safe, secure and robust distributed systems, with capabilities to learn and make informed decisions with respect to their environments. A Boomerang RTDES envisions end-to-end processes operating with potentially sub-millisecond latencies, making it suitable for situations where human control would be insufficient. Through this project, new courses on real-time computing and new research opportunities for undergraduate students are being developed. The project is also working with a number of outreach efforts on the Boston University campus such as the Drone Racing Club, BU BUILDS, SheHacks, and the Spark! Initiative to increase student interest in computing and entrepreneurship. 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.

View original record on NSF Award Search →