ABSTRACT
Today, thanks to the advances in numerical processing and communications, more and more functionalities are embedded into distributed components with the charge for providing the right access to these services. Complex systems are then seen like a collection of interacting subsystems embedding control and estimation algorithms. The inherent “modularity” concept behind this approach is the key answer to the increasing complexity of the systems and this has led to the definition of new models and languages for the formal specification of the components [1]. In this chapter, we are more particularly interested in “intelligent sensors,” components associating computing and communication devices with sensing functions [2]. In order to reduce the complexity, the design of an intelligent sensor requires the necessity to provide a model of the sensor at a high level of abstraction of the implementation. The disparity of the knowledge encapsulated inside the instrument renders the modeling process very sensitive to the modeling strategy adopted and to the models used. A real-life component like the intelligent instrument usually involves the cooperation of three kinds of programs [3]:
• A level of data management to perform transformational tasks. • One or more reactive kernels to compute the outputs from the logical inputs, selecting the suitable
reaction (computations and output emissions) to incoming inputs. • Some interfaces with the environment to acquire the inputs and processes the outputs. This level
includes interrupt management, input reading from sensors and conversion between logical and physical inputs/outputs. Communication with the other components of the system will also be managed at this level.
