Concept and Objectives

Smart Systems represent a broad class of systems that can be defined as intelligent, miniaturized devices incorporating functionalities like sensing, actuation, and control; they are usually energy-autonomous and ubiquitously connected. In order to support these functions, they must include sophisticated and heterogeneous components and subsystems such as:
Application-specific sensors and actuators, multiple power sources and storage devices, intelligence in the form of power management, baseband computation, digital signal processing, power actuators, and subsystems for various types of wireless connectivity (see the figure below).



Smart components and subsystems are clearly developed and produced with very different technologies and materials. Then, the challenge in the realization of such Smart Systems goes beyond the design of the individual components and subsystems (an already difficult task by itself), but rather in accommodating a multitude of functionalities, technologies, and materials; as such, they involve solving problems of different nature.
The widely acknowledged keyword in Smart Systems design is therefore integration. There are essentially two dimensions of integration that represent the main obstacle towards mainstream design of Smart Systems: Technological and methodological. As already experienced in other domains (e.g., digital and analog design), a solution has been found first for the technological issues. Advanced packaging technologies such as System-in-Package (SiP) and chip stacking (3D IC) with through-silicon vias allow today manufacturers to
package all this functionality more densely, combining the various domains depicted in the figure above in a single package. SiP technology works nicely because it allows merging components and subsystems with different processes, and mixed technologies using the stateof-the-arts advanced IC packaging technologies with minor impact on the IC chip design flow.

Design methodologies, however, are falling behind: Current Smart System design approaches use separate design tools and ad hoc methods for transferring the non-digital domain to that of IC design and verification tools, which are more consolidated and fully automated. This solution is clearly sub-optimal and cannot respond to challenges such as time-to-market and request of advanced sensing functionalities. A big step towards effective large-scale design
of Smart Systems would be that of changing the design of such systems from an expert methodology to a mainstream (automated, integrated, reliable, and repeatable) design methodology, so that design costs are reduced,
time-to-market is shortened, design of the various domains is no longer confined to teams of specialists inside IDMs and system miniaturization can be achieved with limited risks.
This objective can be met by defining and implementing a structured design approach that explicitly accounts for integration as a specific constraint, thus minimizing manual hand-off.
The ability of exchanging a wide range of complex design parameters between components and subsystems from different technologies, packages, and architectural templates in a holistic co-design framework is an extraordinary challenge, which requires closing several technical and cultural gaps by means a multidisciplinary approach. The task of constructing a flexible modelling, simulation, design and integration software platform for miniaturized
Smart Systems is thus very ambitious and requires the investment of specific know-how, the availability of human and financial resources and the effort of a large team of dedicated and motivated researchers, scientists and engineers with multidisciplinary and complementary competences.
The first objective of the SMAC proposal is the development of innovative design methods and tools for next generation’s Smart Systems. More specifically, the SMAC vision is that of creating a co-design and integration environment (the SMAC Platform), which enables multi-physics, multi-layer, multi-scale and multi-domain full Smart System simulation and optimization. The co-simulation and co-design environment will be aware of the essential features of the constituent subsystems to be integrated.
The second objective of the project, instrumental to the implementation of the platform, is the development of modelling and design techniques, methods and tools that, when added to the platform, will allow multi-domain simulation and optimization at various levels of abstraction and across different technology domains. This is key for enabling the design of complex Smart Systems using a top-down approach from system architecture down to implementation and integration of heterogeneous functions and technologies.