Descrizione del progetto
Embedded Systems Design
To develop a unique paradigm for the design of embedded control systems that will reduce the gap between control and circuit worlds and reduce the costs in automotive products
Embedded integrated circuits (ICs) are a larger market segment than embedded software and embedded boards. A functionality of paramount importance in embedded systems is their capability to connect with various sensors, actuators, and human interfaces to provide autonomous intelligence in the physical system they are embedded into, such as mobile phones, digital cameras, automobiles, printers, pacemakers, etc. This embedded control functionality is the core element for the combination of the embedded and embedding system to behave properly. As such, highly pervasive and embedded control systems are enablers for much of the high-tech infrastructure that facilitates the lives of many citizens and contributes to their prosperity. Consumer products in our homes, professional equipment, health care, and automotive and transportation systems all heavily rely on embedded control technologies.
In spite of the importance of embedded control in daily life, there is a surprising lack of methods to design and deploy embedded control systems in a systematic and efficient manner. There is an abundance of methods for the separate design of the control algorithms eventually embedded into a hardware platform, or to create a suitable hardware platform for the implementation of a given function. However, an automated and integrated design flow from mathematical models of the embedding physical system to electronic circuits is not available at present.
The MOBY-DIC project will research and develop a unique paradigm and supporting tool chain for the design of embedded control systems, with the aim of reducing both design costs and time to market, thus helping to provide European manufacturers a competitive advantage. Such reductions are dependent on the characteristics of the final product, but in average a 30%-50% could be achieved for a general purpose case.
A key functionality of embedded systems is the ability to react dynamically to changing conditions and guarantee desired behaviour for the physical system they are embedded into. The design of embedded systems that are capable of such control functionalities must take the interactions between the embedded device and the dynamics of the physical embedding environment into account. The heterogeneity between those parts, the ever growing complexity of the overall system, stricter requirements, and the lack of systematic embedded control system design tools, make the design task very time-consuming, expensive, and error-prone. The standard approach is to decouple the design by splitting the control design and circuit design tasks into two separate phases, often executed by engineers with different backgrounds. The only interaction between the two designs is around specs (such as sampling frequency, memory, flops, etc.), which are possibly negotiated iteratively until a satisfactory design is reached, but more often unidirectionally communicated. The resulting outcome is typically far from optimal.
A coherent and consistent paradigm for the design of embedded control systems is the goal of the MOBY-DIC project. A new methodology and associated tool chain will be developed encompassing in a unique framework the modelling of the physical process, design of the control algorithms, design of embedded circuits, and the assessment of the overall performance properties of the system. MOBY-DIC will achieve such an integrated design flow by developing a core methodology based on "piecewise affine" representations, that at the same time (i) provide a rather flexible structure for control functions, and (ii) are directly mapped into digital architectures of small-size and low-power. The effectiveness of the developed embedded control design approach will be demonstrated by MOBY-DIC on a set of challenging applications arising in the automotive industry.
Campo scientifico
Argomento(i)
Invito a presentare proposte
FP7-ICT-2009-4
Vedi altri progetti per questo bando
Meccanismo di finanziamento
CP - Collaborative project (generic)Contatti coordinatore
Coordinatore
16126 Genova
Italia