The research program is mainly focused on the technological issues related to intelligent systems and their ability to be distributed in an environment to solve complex problems. This central topic stems from the description of intelligent systems as products that use communication channels, electronics, microprocessors and software technologies to accomplish predefined tasks, e.g., control objectives given to an autonomous mechatronic system. Embedded systems are fundamental for intelligent products and can be considered as computing systems dedicated to special purposes. These systems will radically change the future through new products as well as sophisticated product development, a trend that has just been started. Indeed, a lot people don't realise that the embedded computer is by far the most common form of computer in use today; most of the consumer products that surround us, such as vacuum cleaners, washing machines or lawn mowers, are more or less complex and embedded systems. Our city streets, homes and working places are supplied with embedded systems in order to control and regulate lighting, comfort and different protection or safety features. There are two relevant aspects of embedded computers in real life: the necessity of robust and dedicated electronics and the increasing need for communication links between each embedded system. Dedicated electronics is in demand since embedded systems are frequently used in harsh environment and add substantial value to products. Indeed, within the next five years, the share of embedded systems is expected to increase substantially automotive, industrial automation, telecommunications, consumer electronics and health/medical equipment markets. Moreover, the value added to the final product by embedded technologies and software is much higher than the cost of the device itself. For example, in case of a modern car, by 2010 over 35% of its value will be due to embedded electronics. This trend is emphasized in a distributed network of embedded apparatuses, where each computing or sensing system may be regarded as a wireless network node that concurs to the global goal achievement.
In this embedded distributed framework, our research activities include, but not limited to, distributed algorithms, to develop lightweight distributed applications to adapt to time-varying and heterogeneous environments and to correctly collect and fuse information gathered from different sources, general methodologies, for system level architectures, paradigms and methodologies design in order to implement resource intensive algorithms, and modelling, related to the development of power aware devices and protocol design. Furthermore, researches in the field of measurements,devising new type of sensing techniques for physical quantities and emphasizing low power and heterogeneous sensors integration and fusion, and of microsensors and MEMS, for highly integrated network nodes, are also deeply investigated. Finally, prototyping for performance assessment using system prototyping-based on resource constrained-platforms will be an additional research field.
Due to the complexity of the distributed embedded systems considered in the research program, research areas cover a large range of fields. Although the main interest remain the embedded electronics and computing systems, great attention is devoted to wireless sensor networks and the communication issues among the nodes of the network. System design methodologies and real-time operating systems are also investigated to implement an efficient software architecture. Machine learning is instead considered for sensor fusion and estimation of the quantities of interest. The efficiency of the solutions proposed is usually proved by simulation and experiments using tools that come from measurement and performance evaluation theory. Due to the nature of the overall system that must deal with continuous quantities and discrete time events (like the node discrete time clock or the communication instants), the loop closure between perception and action is obtained using hybrid control system theory.
Since wireless sensor networks and embedded systems constitute the mainstream of the research program, natural application area is more related to the industrial and environmental monitoring and to the video surveillance. Moreover, domotics and building automation represent other application areas of great interest.
|Luca Abeni||Luigi Palopoli||Roberto Passerone|
|Paola Quaglia||Roberto Sebastiani|
Further InformationResearch Program's Technical reports
Research Program's Published papers