Rafael Priego

Flexibility Support for Homecare Applications Based on Models and Multi-Agent Technology

In developed countries, public health systems are under pressure due to the increasing percentage of population over 65. In this context, homecare based on ambient intelligence technology seems to be a suitable solution to allow elderly people to continue to enjoy the comforts of home and help optimize medical resources. Thus, current technological developments make it possible to build complex homecare applications that demand, among others, flexibility mechanisms for being able to evolve as context does (adaptability), as well as avoiding service disruptions in the case of node failure (availability). The solution proposed in this paper copes with these flexibility requirements through the whole life-cycle of the target applications: from design phase to runtime. The proposed domain modeling approach allows medical staff to design customized applications, taking into account the adaptability needs. It also guides software developers during system implementation. The application execution is managed by a multi-agent based middleware, making it possible to meet adaptation requirements, assuring at the same time the availability of the system even for stateful applications.

SCA extensions to support safety critical distributed embedded systems

Component Based Software Engineering (CBSE) is being increasingly applied in the distributed embedded systems (DES) domain as long as these systems are getting more and more complex in terms of flexibility, dynamism or heterogeneity. Besides that, safety critical systems must cope with the fulfillment of safety requirements and certification standards. This factor increases considerably the development cost of safety distributed embedded systems, even more if they must cope with flexibility, dynamism and heterogeneity. This paper focuses on the distribution aspects of such systems, and more specifically on safe communication channels for safety critical distributed systems. The proposed approach describes a certifiable general purpose safety communication layer that could be reused in different systems, thereby reducing the cost of system development and certification.

Low-cost platforms used in Control Education: An educational case study

Abstract Educational methodologies are continuously improving in Control courses by using high functioning and low-cost platforms. This paper presents a brief description of the most widely used platforms and how they are selected for inclusion in university subjects, specifically in Control Engineering Education. In our experience, such a selection has to be made based on two criteria: The particular concern of the teacher to include more didactic contents in courses and subjects, and the continuously growing experience of students in relation to every low-cost platform. We hereby present a case study of the LEGO(R) MINDSTORM(R) NXT and Arduino platform.

On applying MDE for generating reconfigurable automation systems

The continuous changes on the market and customer demands have forced modern automation systems to provide more strict Quality of Service (QoS) requirements such as load balance or full availability of the control system, among other characteristics. In order to cope with these demanding requirements, modern automation systems have started to incorporate novel design approaches which include reconfiguration mechanisms. Following this trend, this paper presents the use of MDE techniques and technologies, to help in the definition of the control system of reconfigurable automation plants. Directly from the model-based design, a model-based tool is derived that provides support for the generation of the elements that compose the reconfigurable control system.

A model-based failure recovery approach for automated production systems combining SysML and industrial standards

This work presents a failure recovery approach for foreseen failures of automated production systems to minimize the downtime of a system due to stoppages. In contrast to the common practice of implementing field control software, we suggest the use of operation states with pre- and postconditions. A set of operation states forms an operation state machine, whereby several operation state machines are used in a hierarchical manner in order to control and observe the process. The meta-model of the Systems Modeling Language (SysML) is extended to combine operation state machines with OMAC State Machines. By dividing the failure detection from the process controller the necessary flexibility is given to adapt this approach to different packaging machines.

A Model-Based Approach for Achieving Available Automation Systems

Abstract Nowadays automation systems are required to be flexible in order to cope with the ever changing requirements of the applications in terms of complexity, extensibility or dynamism. The use of reconfiguration techniques helps to meet these demanding requirements, but at the same time increase the complexity of the systems. This paper presents a supervisor architecture that allows maintaining the availability of a control system in spite of failures such as a node crash, by means of reconfigurations at the control level. To do this, the advantages that Model-Driven Development technology provides have been explored.

Reconfiguration architecture for updates of automation systems during operation

In recent years automation systems have been required to provide characteristics of flexibility and efficiency in order to cope with the fast changes of the market and consumer demands. To deal with these requirements, the industry has started to implement reconfiguration techniques in order to allow the automation system to reschedule the production, recover from faulty situations and to update the controller at run time, among others. Previous works of authors have been aimed at a model-based definition of the plant components and operations, related to a product. Additionally, automatic code generation has been proposed to obtain an optimal control sequence. This paper extends these previous works with an architecture that generates and updates the software architecture of a Programmable Logic Controller during runtime, in response to changes on the plant layout, the product to be manufactured or the control hardware.

Modeling techniques as applied to generating tool-independent automation projects

Abstract The Industrie 4.0 initiative emphasizes the consistent digitalization and linking of all productive units, including the definition of industrial control systems. This work proposes a model-based approach to develop Industrial Process Measurement and Control Systems. It uses modern software technologies but software complexity is wrapped using the notation, syntax and semantics of the field. It is UML modeling tool independent, due to the definition of domain profiles. It is also PLC programming tool independent, as the software architecture is generated following the PLCopen XML standard.

Agent-based reconfiguration at controller level

Current automation systems need to be able to cope with the continuously changing requirements of the applications in terms of complexity, extensibility and dynamism. To match these challenging requirements, modern automation systems need to incorporate new software engineering techniques. These techniques must provide to the automation systems additional abilities such as diagnosis and reconfiguration. This paper describes an approach for the design of a Multi-Agent System (MAS) based supervisory architecture towards the design of reconfigurable automation systems; the reconfigured system will meet Quality of Service (QoS) requirements. In particular, in this paper the core of the supervisory architecture is presented. It is based on refreshing states and moving functionalities among the Programmable Logic Controllers (PLCs). Besides, the agents negotiate if a PLC failure is detected.

PLC based Supervision of Industrial Switched Processes

Abstract This paper presents a methodology for the design of supervisors for mode switching systems, one class of hybrid system that frequently appears in industrial process applications. The classical continuous control methods are based on designing controllers that are only valid for an operation point for which are designed. The switching among different controllers can optimize transient response and meet multiple objectives. The main goal of this work is to automatically generate the code of the regulation level loops supervision which will be executed on a Programmable Logic Controller (PLC) and following the guidelines for software development provided by the IEC-61131 standard. The tool developed for achieving this automatic code generation processes information about the process operation, the functional specifications and the control algorithms used in each mode. This information is captured from the different view models of the overall control system and through a user interface.