María de los Ángeles Cavia Soto

A Semantic Web Services-based approach for production systems control

Recent research projects have experimented with controlling production system equipment through web service interfaces. However, orchestrating these web services to accomplish a complicated production task can be difficult because the states of the production equipment and the set of available web services may change. This paper proposes an approach in which the production equipment and their states are modelled with an ontology, and the semantic model is dynamically analyzed to determine which action should next be taken in the production process. Similarly, the available web services are described with semantic models, and the descriptions are retrieved dynamically to find the web services capable of performing the required actions. While Web Ontology Language, OWL, is used in describing equipment statuses, the OWL-S ontology, which is based on OWL, is used for describing web services. Once the semantic service descriptions have been analyzed to find the appropriate web services, the services are invoked using their syntactic WSDL descriptions.

Runtime Modeling of Flow for Dynamic Deadlock-free Scheduling in Service-oriented Factory Automation Systems

Abstract Changes in equipment and production demand cannot be predicted at the design stage. Therefore, decision taking mechanisms must rely on real time information collected from the shop floor. To perform scheduling and routing optimization, not only modifications in values of parameters of interest, but also in the flow itself must be accounted for. This paper addresses this problem and proposes a method to formally model, at runtime, the flow within a service-oriented manufacturing line. The resulting representation assists deadlock free dynamic scheduling of the system.

Messaging Performance of FIPA Interaction Protocols in Networked Embedded Controllers

Agent-based technologies in production control systems could facilitate seamless reconfiguration and integration of mechatronic devices/modules into systems. Advances in embedded controllers which are continuously improving computational capabilities allow for software modularization and distribution of decisions. Agent platforms running on embedded controllers could hide the complexity of bootstrap and communication. Therefore, it is important to investigate the messaging performance of the agents whose main motivation is the resource allocation in manufacturing systems (i.e., conveyor system). The tests were implemented using the FIPA-compliant JADE-LEAP agent platform. Agent containers were distributed through networked embedded controllers, and agents were communicating using request and contract-net FIPA interaction protocols. The test scenarios are organized in intercontainer and intracontainer communications. The work shows the messaging performance for the different test scenarios using both interaction protocols.

Planning-based semantic web service composition in factory automation

WSDL annotation conventions facilitate semantic web service description derivation.A software agent composes semantic web services to achieve production goals.Graphical user interfaces visualize domain model and workflow execution statuses.Preprocessing algorithms neutralize syntactic restrictions imposed by AI planners. The Service Oriented Architecture (SOA) paradigm enables production systems to be composed of web services. In an SOA-based production system, the individual production devices provide web service interfaces that encapsulate the behavior of the devices and abstract the implementation details. Such a service-oriented approach makes it possible to apply web service orchestration technologies in the development of production workflow descriptions. While manual formulation of production workflows tends to require considerable effort from domain experts, semantic web service descriptions enable computer algorithms to automatically generate the appropriate web service orchestrations. Such algorithms realize AI planning and employ semantic web service descriptions in determining the workflows required to achieve the production goals desired. In addition, the algorithms can automatically adapt the workflows to unexpected changes in the goals pursued and the production devices available.

Cloud computing as a facilitator for web service composition in factory automation

Cloud computing is an information technology paradigm enabling companies to sell computing resources more dynamically. Software and hardware are now commodities leased on demand. Because computer systems leased from a cloud service provider, virtual machines, are typically connected to internet, they can host web services, which are frequently components of service oriented architectures (SOAs). Such architectures have recently been adopted in factory automation, as they allow systems to reach high levels of decentralization and loose-coupling. SOA-based Factory automation systems combine physical production equipment with web services that belong to the information processing (cyber) domain, and they are therefore highly cyber-physical. When some of the services are deployed on cloud resources, SOA-based factory automation systems can be classified cloud-based cyber-physical systems. Each service in such a system is typically able to perform rather simple, atomic operations, whereas achievement of complex goals requires that the services be composed to collaboratively carry out workflows. This article investigates the use of cloud resources in automatic service workflow composition. To facilitate the acquisition and utilization of cloud resources, a system of two specialized web services is proposed. The system includes a web service that dynamically deploys virtual machines to carry out planning processes, thereby exhibiting artificial intelligence. Finally, this paper demonstrates the integration of the system with a previously proposed semantic web service composition framework.

A modelling approach to formally represent service orchestration

This paper presents a methodology to formally model service orchestration. The methodology is based on a set of selected flow descriptors, which are expressed in a Petri net derived formalism. The chosen descriptors cover concurrency, synchronisation, multithreading, looping and conditional execution. The modelling of temporal and logical conditions arising in conjunction with looping constructs is dedicated a satellite set of models. The formal definition of the translation procedure for the structured activities of the main set is given. The means of calculating the relationship between the timing constraints of the constructs used to specify the orchestration and the intermediate constraints of the composing services are studied. The paper concludes with a presentation of the industrial demonstrator used in this research and a discussion about the advantages and disadvantages of the presented modelling approach, at verification stage.

An event-based service-oriented infrastructure for reconfigurable manufacturing systems

Volatile markets require that associated manufacturing systems cope with a variety of products and frequent changes in capabilities. The required paradigm is that of Reconfigurable Manufacturing Systems, which are designed for gracefully assimilating changes in functionality. This paper proposes a middleware infrastructure that facilitates the process of integrating and (re)configuring manufacturing systems. The middleware follows a layered approach, in which an information infrastructure is implemented over an interaction infrastructure. The interaction infrastructure encapsulates the interaction primitives required to manage distribution in a loosely coupled way. A service-oriented approach is used for discovering deployed elements. A message-oriented approach is used to asynchronously propagate invocations and events. The information infrastructure encapsulates functionality to facilitate the invocation, using an upper ontology for Semantic Web Services in conjunction with domain ontologies in order to semantically model processes, products, equipment, the world and the relations between them. Three use cases extracted from a pilot assembly line are employed to illustrate the use of the infrastructure.

Differences using measured and calculated solar radiation in order to estimate the temperature of the conductor in overhead lines

This work was supported by the Spanish Government under the R+D initiative INNPACTO with reference IPT-2011-1447-920000 and Spanish R+D initiative with reference ENE2013-42720-R. The authors would also like to acknowledge Viesgo for its support.