Semantic Web Enabled Network of Maintenance Services for

Semantic Web Enabled Network of Maintenance Services for Smart Devices “Industrial Ontologies” Group Tekes Project Proposal http: //www. cs. jyu. fi/ai/Metso_Maintenance. ppt Agora Center, University of Jyväskylä, March 2003

Our Team: “Industrial Ontologies” Group § Head: l vagan@it. jyu. fi Agora Center, University of Jyväskylä Vagan Terziyan § Researchers: l l l Oleksandr Kononenko Andriy Zharko Oleksiy Khriyenko § Supervisor and Consultant from Metso: l Jouni Pyotsia “Industrial Ontologies” Group: http: //www. cs. jyu. fi/ai/Onto. Group/index. html

Emerging Semantic Web § “Knowledge is an important productivity factor” § However to make your knowledge to be really such you should consider managing it based on emerging Semantic Web Technology § Then it would be possible to take better care of your businesses, products, services, processes, etc. using automatically collected and integrated experience from different heterogeneous distributed sources worldwide § This makes possible also to make your own knowledge and experience reusable, shared and permanently beneficial

Enterprise Integration Technologies § Web Service Technology (SOAP, WSDL and UDDI); § Enterprise Integration (Enterprise Application Integration and E-Commerce in form of Business-to-Business Integration as well as Business-to-Consumer); § Semantic Web Technology (ontology languages). The promise is that Web Service Technology in conjunction with Semantic Web Technology (“Semantic Web Services”) will make Enterprise Integration dynamically possible for all types and sizes of enterprises compared to the “traditional” technologies

Semantic Web § “The Semantic Web is a vision: the idea of having data on the Web defined and linked in a way that it can be used by machines not just for display purposes, but for automation, integration and reuse of data across various applications”. http: //www. w 3 c. org/2001/SW

Ontological Vision of Semantic Web needs ontologies An ontology is § document or file that formally and in a standardized way defines the hierarchy of classes within the domain, semantic relations among terms and inference rules Use of ontologies: § Sharing semantics of your data across distributed applications

Knowledge Management based on Semantic Web concepts A commitment to a common ontology is a guarantee of a consistency and thus possibility of data (and knowledge) sharing It seems feasible to use standards of the Semantic Web research community for the development of next-generation information systems based on ontology-driven knowledge management, e. g. : § Intelligent process automation systems § Intelligent condition monitoring systems § Decision support systems (embedded AI) § Intelligent maintenance systems and services § …

Project Primer Goal § The primer goal is to study and implement the benefits of the: l l l Semantic Web (interoperability based on ontological support and semantic annotations), Intelligent Web Services (modelling, automated discovery and integration), and (Multi)Agent technologies (agents communication, coordination and mobility) § … to improve the performance of the Field Device Management Process by launching a network of distributed intelligent maintenance services.

Pilot Implementation Goal . § More specifically the goal is to develop: l l l a prototype of a global intelligent diagnostics and maintenance support system, an appropriate multiagent support for it, ontological support for it, pilot prototype implementation, case study.

Global vision: agents in action Agents acting as service components in the Maintenance Service Network have ability to learn during work improving services’ performance. New vision assumes a Maintenance Services Network of smart-devices and Maintenance Service Centers, in which maintenance experience is accumulated independently by agents of each Maintenance Center with a possibility to be integrated together when needed. Smart-devices are becoming users of provided maintenance services.

Challenge 1: Service Users are devices § The class of service requestors is extended with new group of service users – smart devices. § We add semantic-enabled descriptions of services to facilitate: l automated discovery and use of services by smartdevices; automated integration of services; l communication between heterogeneous services. l

Distributed knowledge I’m competent in domain 1. . I have a problem from domain 23. . Who can help? I’m competent in domain 2. . I’m competent in domain N. . As a result of independent maintenance experience accumulation by service components (agents) every Maintenance Service Center in the net provides specific set of service components. When a problem arises maintenance service components with the most relevant knowledge for that case might be found in the net.

Internal and External Agent Platforms Service Platform Environment where service components perform: • Condition monitoring • Maintenance activities Maintenance Platform Environment to run Maintenance Services, contains a set of expert-agents both in maintenance and diagnostics. Agents are “service components” Based on the online diagnostics, a service component-agent, selected for the specific faulty or emergency situation, can be moved to the service platform to help the host agent to manage it and to carry out the predictive maintenance activities.

Challenge 2: Two Types of Service Platforms § Service Platform is an environment for running services and hosting service components (agents). § Services can be provided either locally, i. e. by embedding them to smart-device internal platform, or remotely by querying them from a Web-based external platform. § External service can be queried either from Web-based external platform or from another internal platform. § External Web service platforms provide more rich services since they are used by many clients and quality of services can be permanently improved according to growing experience. § Various interactions between service platforms (internal-internal, internal-external, external-external) can be organized as a P 2 Plike network.

Internal Platform Recovery components Diagnostic components Management component Field Agent Wrapper component Field Agent – device-dependent embedded condition monitoring component (e. g. Field. Browser); Wrapper component – for integration with device-dependent (software and hardware) resources, acts as a semantic adaptor, mediator between semantic-enabled and traditional parts of service infrastructure; Management components – for management of maintenance activities and distributed resource allocation; Diagnostic components – for online discovery of problems within a device based on its state parameters and ontology-based classification of these problems (component is mobile agent); Recovery components – for automatic planning and performing appropriate maintenance activities for a discovered diagnosis (component is mobile agent).

External Platform Recovery components Diagnostic components Management component, Diagnostic components, Recovery components – service components of Maintenance Service Center. There is similar service components set as in the Internal System structure, but these components have more rich “experience” and abilities to solve problems.

Agents in Semantic Web 1. “I feel bad, pressure more than 200, headache, … Who can advise what to do ? “ 3. “Wait a bit, I will give you some pills” Agents in Semantic Web supposed to understand each other because they will share common standard, platform, ontology and language 2. “ I think you should stop drink beer for a while “ 4. “Never had such experience. No idea what to do”

The Challenge: Global Understanding e. Nvironment (GUN) GUN How to make entities from our physical world to understand each other when necessary ?

GUN Concept 2. “I have some pills for you” 1. “I feel bad, temperature 40, pain in stomach, … Who can advise what to do ? “ Entities will interoperate through Onto. Adapters, which are “supplements” of these entities up to Semantic Web enabled agents

Semantic Web: Before GUN Semantic Web Applications Semantic Web applications “understand”, (re)use, share, integrate, etc. Semantic Web resources Semantic Web Resources

GUN Concept: All GUN resources “understand” each other Real World Objects + + Onto. Adapters = Real World objects = GUN Resources GUN Onto. Adapters

Maintenance Services Organizing the maintenance … • Service 1: Remote diagnostic • Service 2: Recovery and predictive maintenance • Service 3: Preventive inspection • Service 4: Emergency service • Service 5: Human resource execution • …

Remote diagnostics scenario parameters diagnosis Agent with knowledge • Alarm situation is locally detected however Internal Maintenance Platform (IMP) is not able to classify it as certain diagnosis. Thus IMP sends request with parameters to an External Maintenance Platform (EMP). • As a result, EMP sends discovered diagnosis back to the IMP. • If similar request for diagnosis is sent often enough, then it is considered to send appropriate diagnostic service component (mobile agent) from EMP, to operate locally at the IMP.

Challenge 3: Service Components are Autonomous Intelligent Agents § Service components are mobile; § Service components are able to learn; § Service components are Semantic Web enabled

Maintenance Service Network Service components are certified. Certification system is a basis for guaranteed quality of maintenance services. All interactions in the Maintenance Network are performed between Management Service components Requirements for Management Service component: • Check of request correspondence to available local services, based on profile of MC. • Request to other components of the network, in case if request can’t be satisfied. • Enabling peer-to-peer semantic search in the Maintenance Service Network

Service management MC Profile is a file, that contains information about: - what type of maintenance activities MSC provides; - what level of quality it’s gained during certification; - economical aspects (cost). Since we have independent services in distributed environment, the peep-to-peer concept must be implied on base of Semantic Web (profile web). High-level functions are performed on the base of profile processing. Each Maintenance Service Center has a corresponding profile which describes its services. Profile is created in machine understandable form on a basis of common ontology.

Challenge 4: Semantic P 2 P Concept for Service network Management The concept assumes decentralized management architectures with. centralized ontologies for e. g. : § Service certification management; § Service discovery management; § Service responsibility management; § Quality of Service management; § Trust management; § Privacy and security management. Also transaction management issues related to transportation of mobile. components between platforms should be addressed in this project. Two levels of management are considered: for interactions between. local service platforms of smart-devices (P 2 P network) and for. interactions between service centres on enterprise level.

Subdomain ontologies The following set of subdomain ontologies can be defined: describes device structure, its components and states (for maintenance/control processes) describes breaks and faults classifications, maintenance cases bindings to certain products or components, specification of detection methods, rules, etc. describes maintenance activity classification and prerequisites of use: rules/inference tools to use, etc.

Upper Maintenance Ontology Class of Maintenance Activities Class of Diagnosis Applied-to State of product Class of Resources Subclass-of Restriction Subclass-of Diagnosis Class of Restrictions State Restricted-by Maintenance Activity Product Procedure Requires Resource Subclass-of Class of Standards Subclass-of Specification Standardized-by Standard

Creating ontologies… Class details Classes hierarchy Class properties (“slots”)

RDF in XML RDFS: vocabulary RDF: description

Ontology of Control Valves with Protégé

Our project implementation goals Implement minimal set of maintenance service components (agents) Provide minimal set of necessary information structures and ontologies Design Maintenance Centers Infrastructure Design Pilot Service Platforms

Necessary data for pilot implementation To select some product as a case for implementation and consider different diagnostic cases. What is the relation ’maintenance activity’ – ’equipment state’ What maintenance activities exist? How equipment state is described? What is the relation ’equipment state’ – ’breakage class’ How does ’breakage class’ associate with ’maintenance activity’ What breakage classes exist?

Project Main Objectives 1. Development of upper-ontologies for the maintenance domain 2. Development of samples: (a) an embedded agent-enabled platform and (b) Semantic Web maintenance service for smartdevices 3. Development of P 2 P semantic search techniques in semantic -enabled network of maintenance services 4. Pilot implementation of embedded platform and set of maintenance services Development of ontology for Metso smart-device case 5. Testing of pilot system on the Metso smart-device case

Project Deliverables 1. Requirements to a Maintenance Service Network for Smart. Devices 2. Requirements to possible service components (agents) 3. Requirements to an embedded service platform 4. Requirements to ontology management in a semantic P 2 P network 5. Requirements to maintenance service ontology 6. Scenarios for certification, security, privacy and trust management 7. Service platform specifications and implementation plan 8. Upper-ontologies for smart-devices’ maintenance domain § § Diagnostics ontology § Maintenance activities ontology § 9. Devices ontology Maintenance service ontology Pilot implementation of the Service Platform

Conclusions § Traditional Enterprise Integration technologies are able to address some of maintenance management problems today. However, new technologies like Web Services Technology in combination with Semantic Web and Agent Technologies have the potential to address maintenance needs much better § We have experience and human resources to develop the concept of Distributed Maintenance Network and provide implementation starting from a pilot system and pilot ontologies § Results can be used by co-operating companies: e. g. Metso for management of their field devices based on embedded agent platforms and Web services; Sonera for providing communication infrastructure for embedded agents and launching appropriate Web services for this and also for other cases