Intelligent Instruments: Some Modelling Approaches

In the early eighties, the CIAME was set up by sensor vendors, researchers and users who decided to work together to compare their points of view on sensor and actuator developments. Within this frame, the CRAN (Centre de Recherche en Automatique de Nancy) and the LAIL (Laboratoire d' Automatique et d' Informatique Industrielle de Lille) have initiated fundamental research on intelligent instrumentation in France. The results obtained have been enriched by industrial glances in CIAME working groups. We now recall the distinction done by Kleinschmidt between smart and intelligent, as applied to the so called equipment• In order to clarify the vocabulary, the terminology 'smart instrument' (sensor or actuator) refers to a stand alone instrument possessing the functionalities described in section 2, while the terminology intelligent equipment refers to the equipment being part of an AS (Automation System) and contributing to its 'intelligence'. The research work of the CIAME was essentially concentrated on making a census of services to be offered by intelligent instruments and an attempt to reach a functional definition for these instruments. The results of their reflections have been gathered in two books on the subject3.4. At the same time, a retlection on the evolution of automation production systems had been undertaken. The increase in industrial competitiveness and the demand of customers for consistent production quality, have led control engineers to reorganise automation production systems. This reorganisation has been facilitated due to the services offered by intelligent instruments. Indeed, part of the reorganisation was based on the distribution of intelligence within the APS and an extensive use of fieldbuses. The intelligent instrument functionalities and services coupled with fieldbuses allow a direct connection of sensors, actuators and controllers, creating decentralised control loops to off load central computers of burden control functions. Automation systems conforming to this type of configuration may be called Distributed Intelligence Base Automation Production Systems (DIBAPS). Distributed intelligence and the use of fieldbuses have also allowed us to envisage a better integration of the three distinct islands of an APS (Automation Production System), i.e.: • the Control island • the Maintenance island • the Technical Management island. While in classical APS these islands are managed independently, the use of fieldbuses and intelligent instrument functionalities and services allow a propagation of consistent and coherent information among the three islands, thus tending towards their integration: the CMMS (Control, Maintenance and technical Management Service) concept. The distribution of intelligence in APS through intelligent instruments must first of all call for a standardisation of the functions of these instruments, in order to provide them with the ability to cooperate. One means by which this standardisation can be achieved is through a modelling of intelligent instruments. In this paper, we successively discuss: • Some essential smart instrument functionalities. • Two different approaches for the design of intelligent instruments. • Two models which can be reused in the design of DIBAPS. The application of one of the two modelling approaches to a smart temperature transmitter is given, while the generic character of the last model is illustrated.