Embedded and Hybrid Systems (EHS)

Program Scope

The Embedded and Hybrid Systems program supports research in scientific foundations and systems technology to revolutionize the design and development of embedded systems for applications at all scales.  In the past, embedded systems were typified by simple, cyclic real-time executives, running on 8- or 16-bit processors with only a few thousand bytes of memory.  Although some embedded systems remain at that scale, current trends are to 32-bit processors augmented with digital signal processors or reconfigurable programmable gate arrays, and in some cases to general purpose processors with significant memory for data storage and control.  Applications include avionics, medical, and automotive systems, in which electronics and computation have substantially augmented mechanical functionality.  They include complex systems-of-systems such as the national power grid. The hallmark of an embedded system is its interaction with a physical system, from which its constraints derive.  Embedded systems software and information technology now serve as a key accelerator of progress in computational control and engineered systems.  

The EHS program supports research in all aspects of embedded systems, including such topics as:  software technology for digital devices enabling personal communication and computation; dynamically configurable real-time operating systems services, virtual machines, and middleware; algorithms and systems services to coordinate autonomous sensor nets and ubiquitous computation; foundations and technology for real-time systems; distributed real-time sensing and control; and integration and coordination mechanisms for large-scale, complex, real-time systems and systems-of-systems.  A pervasive theme of the EHS program is the high-confidence integration of real-time and other service guarantees with the coordination requirements of next-generation complex, secure, networked, embedded systems.

Background

Software and information technologies enable increasingly ambitious, often safety-critical, systems (e.g., transportation, manufacturing, medical devices and systems, environmental control, and energy management). Today and in the future, these must include distributed and coordinated embedded systems that demand a high degree of autonomy, dynamic adaptation, and component integration.  Examples are seen in the multimodal sensing and control systems that must integrate the operation of critical subsystems and multi-systems in manufacturing and control, vehicles and transportation systems, telecommunication, and power transmission. Hard real-time operation may be mixed with general and interactive computation, and coordinating subsystem operation is a key function.  Embedded software for these systems typically must integrate interacting system aspects: temporal, spatial, and physical properties and underlying dynamics of the system to be monitored or controlled; the timing and synchrony properties and resource demands of software that controls the system; and the characteristics, resources, and services of the computational platform (both systems software and hardware). For this reason, the EHS program includes a focus on the hybrid discrete and continuous nature of computational problems for this class of systems.  Another major emphasis is on open systems approaches that enable dynamic adaptation and decentralized operation.  Finally, the program encourages research to exploit advanced and emerging computational and sensing technologies, including micro- and nano-systems.  The goal of the EHS program is to create and unify the foundations for managing the interaction of physical and computational systems and to supply the technologies needed for building reliable software- and network-enabled embedded systems.  The program draws on computational aspects of control theory, modeling and simulation, property-preserving software generation and composition, real-time and power-aware design, systems software and middleware, and formal and informal assurance methods. 

The Embedded and Hybrid Systems program seeks both far-reaching innovation and integration in areas such as: 

• Embedded software composition: middleware, virtual machines, and real-time operating systems services for distributed, real-time, embedded systems; device driver synthesis; languages, algorithms, tools, patterns, and component frameworks for embedded systems; programming and composition technology for integrating timing, fault-tolerance, security, data coherence, and other cross-cutting properties into functional software
  
• Resource management and optimization:  models, algorithms, and analyses; compilation and linking techniques, algorithms, frameworks, and tools for allocating, scheduling, and managing computation, communication, and power resources in real-time distributed and embedded systems; quality-of-service mechanisms
  
• Foundations and information technology for software-enabled control:  modeling, design and implementation of software controllers; hybrid discrete-continuous systems; model-based design environments; controller synthesis; integration of digital and supervisory control; control coordination frameworks and protocols
  
• Distributed sensing and control technology: scalable support for embedded sensors and sensor nets; ubiquitous sensing and control; highly innovative strategies and information technology for future supervisory control and data acquisition (SCADA) systems 
  
• High confidence embedded software and systems: foundations, design, implementation, analysis and certification approaches for embedded systems