Available online at www.sciencedirect.com
Procedia CIRP 2 (2012) 22 – 27
1st CIRP Global Web Conference on Interdisciplinary Research in Production Engineering
AutomationML server - A prototype data management system for multi disciplinary production engineering S. Makrisa* and K. Alexopoulosa a Laboratory for manufacutirng Systems and Automation Director Prof. G. Chryssolouris, University of Patras, Greece * Corresponding author. Tel.: +30-261-9970262; fax: +30-2610-997744.E-mail address:
[email protected].
Abstract Production engineering is a complex procedure since distributed engineering teams should collaborate, using heterogeneous IT tools. This work presents the concept and initial implementation of an AutomationML server for collaborative computer-aided production engineering in which Original Equipment Manufacturer and supplier companies interact with the help of a diverse range of design and engineering tools. The server provides collaboration and data services to electromechanical objects via a collection of predefined Web Services. Finally, the development of a web AutomationML editor is discussed and its integration to the proposed AutomationML server is presented.
©2012 2012The Published by Elsevier Ltd. Selection and/orand/or peer-review under responsibility Ir.Wessel Wessel Wits © Authors. Published by Elsevier B.V. Selection peer-review under responsibilityof of Dr. Dr. Ir. W.W. Wits Keywords: AutomationML server; Ramp up; Multi disciplinary engineering
1. Introduction The decreasing lifecycle times of vehicle models as for numerous other products and the increasing number of variants require from automotive industry to design and operate assembly plants and production networks that are completely flexible, capable of switching from producing one model to another to meet fluctuating and diverse demand. A key issue in order to succeed in such diverse, complex and customer driven market-context is to produce on-demand products of high complexity, while improving quality and product-service life cycles. In order to design and setup such flexible production plants, manufacturers have to improve the development process from product design up to production. Production engineering involves the design, control and improvement of manufacturing systems in order to provide the customers with high-quality products, in a timely and cost efficient manner. Product design, process and production planning procedures are supported by advanced simulation tools and models, computer aided design / manufacturing (CAD / CAM) software packages, enterprise resource planning (ERP) and supply chain management (SCM)
systems based on computer-integrated manufacturing (CIM) concepts, along with Product Lifecycle Planning and Management (PLM) methods. Product Lifecycle Management (PLM) is a collaborative product development supply ecosystem that enables manufacturers to manage a product from its early concepts to its retirement. Nevertheless and despite the existence of some digital factory planning models and tools for the different sections of a vehicle production, in today’s manufacturing projects, there is a clear separation of planning and realization phases; while the planning is mainly done by the OEMs themselves, the realization is assigned to production equipment suppliers [1]. In principle, the integration of control systems with CAD / CAM and scheduling systems as well as real-time control based on the distributed networking between sensors and control devices [2] currently constitute key research topics. 2. Collaborative production engineering Production ramp-up is the period between the start of production (SOP) and a stable production with the
2212-8271 © 2012 The Authors. Published by Elsevier B.V. Selection and/or peer-review under responsibility of Dr. Ir. Wessel W. Wits http://dx.doi.org/10.1016/j.procir.2012.05.033
S. Makris and K. Alexopoulos / Procedia CIRP 2 (2012) 22 – 27
targeted output. Ramp-up relevant costs have been quantified being 3-7% of the overall investment of a product launch. During the ramp-up phase 7-15 % of additional work force is needed to manage uncertainties and to ensure the execution of new and undiscovered processes. Further costs are resulting from call backs and are also caused by quality failures during ramp-up. Only in North America the warranty charges for callbacks in the automotive industry are quantified with EUR 10 billion. In the near future the economic success of enterprises will be dominated more and more by the efficiency of their ramp-up processes [3]. Virtual assembly plant is a concept used to reduce the ramp up time by validating the production system within a virtual environment. Its focus falls in the area of Prototyping for Fast Ramp-Up, where current concepts, seem inefficient in a key element: the integrating factor between product design and production (assembly) planning [4]. The aim of the virtual assembly plant is a full-scale development of virtual assemblies (prototypes) capable of being verified via digital simulation, where the rampup phase is shortened since the need for initial physical installation of the production line at the system supplier's site is replaced by its full digital replica. The main problem is that modern assembly systems are faced with the cost-driven demand for faster and more secure ramp-up processes. This goal is underpinned by the rising number of ramp-ups, due to enhanced innovations and increasing market launches of new products and product variants. The complexity and diversity of the different line components in terms of control systems and communication channels requires a lot of time for onsite setup, testing and validation of the assembly equipment. This in turn, is translated in production system downtime and the respective opportunity costs that follow it. Collaborative production engineering is the way that modern enterprises face the issue of ramp up time reduction, involving a wide range of multi disciplinary engineers in the process of production engineering. A production engineering project is an instantiation of the manufacturing systems’ design projects in which the designer must describe a suitable design that can map the manufacturing system variables (e.g. number and type of machines, process parameters, control logic etc) to performance requirements (e.g. product characteristics, production cost, flexibility etc) [5]. Nowadays, integrated production engineering is operated within the entire network of collaborating enterprises, defined as Extended Enterprise (EE), with a long term agreement, existing from supplier to end-use customers. The expansion of the Internet provides the infrastructure by which information can become simultaneously available to all those involved in production engineering,
such as, designers, planners, production managers, and so forth. However, there are problems with the automation of the production engineering in the EE: The generation and the execution of a production engineering project may take a long time and may involve several organizations in different geographical locations. Similarly, monitoring and improving the production plans may be complex and difficult to automate. The commercial Computer-Aided Manufacturing/Engineering (CAM/CAE) systems, used by process designers and planners, can be different. The internet-based manufacturing needs to overcome this heterogeneous software environment. Finally, data inconsistency and data redundancy should be addressed. These problems can be tackled by distributed, adaptable, open and intelligent process planning systems within a collaborative environment. In a geographically dispersed production engineering environment, web-based tools for collaboration are highly relevant but they are not a panacea. They should be complemented by a framework that enables the integration and coordination of engineering activities (product development, process planning, commissioning) and the exchange of data among entities (Original Equipment Manufacturers –OEM-, engineering teams, system integrators etc). In the case of an EE different tier engineers collaborate within the common communication infrastructure, provided by the OEM. The engineering tasks are usually accomplished among the OEM, the main integrator for the complete system, and probably additional engineering service providers. The production engineering projects are typically multidisciplinary and inter-organizational. The current situation in digital production engineering is characterized by a large number of different IT tools, both PC based and main frame applications [6]. The backbone for engineering activities is a PLM. There are tools in support of mechanical design operations, process design and simulation, electrical engineering, OLP (Offline Robots Programming), PLC coding and automation. In [7], they proposed a BPMN based method (Business Process Modeling Notation) and implemented it within a commercial PLM (Product Lifecycle Management) system for a collaborative process planning environment for integrating PLM by CAPP and CAD/CAM tools. However, equipment automation aspects are not considered, which add additional requirements to the interoperability aspects within the extended enterprise and in their approach, there is not a service oriented architecture framework to support collaborative data storage/retrieval and communication issues. In [8], they focus on the use of a serviceoriented-architecture (SOA), namely web services, intelligent software agents to support integrated
23
24
S. Makris and K. Alexopoulos / Procedia CIRP 2 (2012) 22 – 27
engineering activities. The pilot case execution has indicated a significant man-hours saving, due to the coordination of engineering activities and automation. But their approach has used a proprietary implementation to bind business process execution with SOA architecture but not some standard approach such as BPEL. Other approaches that demonstrate the use of an agent based framework for the integration of heterogeneous software tools, even non-engineering application fields, can be found in the literature (e.g. [9], [10]). In [11], the implementation of a multiagent system in the production engineering is discussed, aiming at an emergent behavior and an adequate structural and organizational model of the real-world system. Communication protocols can be autonomously generated, refined or adapted by the agents. The approach uses basic concepts of machine learning for integrated conflict management and open adaptive communication. By this, a highly complex communication structure emerges during the runtime. Finally, in [12] a web-based workflow management system for integrated production engineering along with the first implementation of a software system that manages AutomationML production engineering data. However, in [12] only the basic components of the AutomationML server are presented. The work presented in this paper complements the work presented in [12] by describing significant details of the proposed Web-Services API structure and also by providing an example of external application integration to the AutomationML server.
3. System architecture and implementation The work in this paper, focuses on typical production engineering scenarios, where an OEM communicates with geographical dispersed engineering teams in order to define the equipment (e.g. robots, grippers, PLC devices, weld stations, slide-in carts etc) required for production. Furthermore, it is common that the engineering enterprise selected, requires the planning of its engineering activities in a collaborative way, in order for the engineers' knowledge and the relevant engineering tools to be incorporated with efficiency and robustness. To achieve these objectives, it is required that all collaborative activities and data flows (product data, manufacturing process plans, engineering software tools etc.) be identified, and that they should be integrated into a management infrastructure that allows all the communication and efficient data exchange. In this work, the AutomationML server has been developed in order to address collaborative production engineering requirements. The AutomationML Server is an AutomationML data management prototype that
serves AutomationML objects for production engineering projects via a set of Web Services. AutomationML (Automation Markup Language) open standard [13] which is an XML data format for the exchange of plant engineering information is used for the data exchange among heterogeneous production engineering tools such as mechanical plant engineering, electrical design, HMI development, PLC, robot control etc (see Fig. 1). The main objective of AutomationML is to increase the interoperability between different engineering tools [14]. If there is a need to support cooperating partners who share data in different formats, from different engineering tools, then AutomationML may be employed as the means for data exchange. Process Engineering Web Services S
Mechanical Engineering Web S Services
Electrical Engineering
Robot Engineering
PLC Programming
Web Services Se
Web Services S
Web Se Services
http://internet t
Web e S Services vc
Web b Services
Workflow Management
AutomationML Server
Fig. 1: AutomationML server within the heterogeneous production engineering environment
AutomationML server is a back-end server solution that may support cost effective management of AutomationML production engineering data, both across the production engineering lifecycle and across system and company borders. The architecture of the AutomationML server is shown in Fig. 2. The functionality of the server is provided through a set of well defined Web Services. The implementation of the web services is performed in the "internal" application tier which is also responsible for the permanent storage of the data into an Oracle 11g relational database that has been selected for data persistency. The Web- Services API has been structured in separate modules, each targeting at a specific information scope. There are general purpose modules and business ones. The modules and the services provided are described hereafter: x
General purpose modules: This set of modules provides functionalities for managing a project created in the AutomationML server. o AccessManagement: The services of this module: a) enable a client application to
25
S. Makris and K. Alexopoulos / Procedia CIRP 2 (2012) 22 – 27
o
o
x
authenticate itself into the AML server and acquire a session, b) enable users management (add/remove/change credentials) and c) grant access of users to AutomationML based production engineering projects. ProjectManagement: The objective of this module is to provide functionalities for managing an AutomationML project’s lifecycle. QueryManagement: This module provides search queries to the AutomationML service database.
Business targeted modules o AMLDocumentManagement: Provides services for uploading and downloading AutomationML files and referenced documents. o AMLElementManagement: Provides functionalities for managing AutomationML elements. o AMLDocumentExternalFilesManagement: The services of this module enable the management of external files to an AutomationML document. o AMLSchemeMaintenanceManagement: Provide services for maintaining the AutomationML scheme supported by the server in a "per project" basis. It maintains the AutomationML metadata of the CAEX scheme. o CAEXElementManagement: This module provide functionalities related to InternalElements of a CAEX file. o AMLEngineProxy: The objective of this module is to publish as Web Services the functionality of the AutomationML engine. The functionality provided in this module is independent of the AutomationML server functionality. However in order for a user to able to use the functionalities authorization is required.
An example the web-services one AutomationML server module is presented. AMLElementManager module services: - addCOLLADAReferenceAttribute(long projectId,String elementGUID,AttributeType colladaRef); - addExternalInterface(long projectId,String elementGUID,String name, InterfaceClassType interfaceElem); - addFrameAttribute(long projectId,String elementGUID,AttributeType amlFrame);
-
addLogicReferenceAttribute(long projectId,String elementGUID,AttributeType logicRef); removeCOLLADAReferenceAttribute(long projectId,String elementGUID); removeExternalInterface(long projectId,String elementGUID,String name); removeFrameAttribute(long projectId,String elementGUID); removeLogicReferenceAttribute(long projectId,String elementGUID); updateCOLLADAReferenceAttribute(long projectId,String elementGUID,AttributeType colladaRef); WSDL
Web Service Integration API
J2EE
Web Tier AML Direct Web Services
Client Applications A Client
AML Web Services
AML Editor
Application Tier AutomationML Services Logic
Enterprise Application Integration API
Workflow
Application Management Access Manager
Session Manager
Database Manager
AutomationML Data Model CAEX
AutomationML
Data Tier Database
File System
Fig. 2: AutomationML Server architecture
4. Application integration example In this paragraph, there is a presentation of the way that the proposed framework AutomationML data management server can be integrated with an application that utilises AutomationML data. For that purpose The Web AutomationML Editor has been developed. Web AutomationML Editor is a tool, based on Java applet technology that enables editing of AutomationML files from directly within a web browser. This tool offers the following AutomationML capabilities to the end-user: x Graphical User Interface (GUI) for editing AutomationML objects in a tree structure. The GUI enables operations on AutomationML Internal Element objects such as: add/remove an Internal Element, add/remove Attribute (COLLADA, Frame, Logic), add/remove references to External Interface (such as COLLADA or OpenPLC), change parameters of AutomationML objects (e.g. name). Provides operations on AutomationML Hierarchy. x
AutomationML data loading/storing: The tool supports two modes for data sources a) typical data file access from a local directory; in this mode the Web AutomationML editor loads data from the local disk and saves changes back to the disk and b) accessing data through Web-Services through from AutomationML server. In this mode the application
S. Makris and K. Alexopoulos / Procedia CIRP 2 (2012) 22 – 27
26
acts as an AutomationML client, the AutomationML data is permanently stored on the AutomationML server. The data loading/storing operation mode is configured during the initialization phase of the tool. x
Basic COLLADA viewer for viewing COLLADA geometry files. The viewer support basic navigation operations (rotation, pan and zoom). The COLLADA geometry is displayed in the Viewer by the time an element with relevant geometry information is selected in the tree structure pane.
Fig. 3: Web AutomationML editor
The screenshot of Fig. 3, represents the User's Interface of the Web Automation ML editor. On the left side there is a tree structure presenting the AutomationML data objects and on the right side the COLLADA viewer. The Web AutomationML editor applet has been integrated in a proprietary workflow management tool [12] in order to allow to the users of tasks of the workflow that require editing AutomationML data to be able to execute AutomationML editing actions within the workflow application. Different use case scenarios, that involved setting up and AutomationML project and editing different AutomationML attributes were applied using the web editor and the server. It has been shown that the proposed architecture provides the ability to efficiently share AutomationML data among different users for different steps of a production engineering project workflow. 5. Conclusions and outlook The paper discussed the implementation of an AutomationML data management system capable of managing the multi disciplinary the production engineering process. Modern information technologies were utilized to implement the proposed concept. The
AutomationML server may act as a space for any application that requires management (loading, storing, versioning, accessing, editing) of AutomationML data. Through its open definition of Web Service it can be integrated in a "Plug & Play" approach with a wide set of engineering applications that need a "place" to store and "share" AutomationML data. The comprehensive set of web services allows rapid development of additional clients. The AutomationML server leverages the AutomationML from a file based data exchange format for engineering tools to a federation of engineering data though the web services framework. Depending on the systems from which data is sourced, setup and implementation of AutomationML Server, is expected to be able to be completed in a few weeks. Unlike the traditional approach of many PDM and PLM systems, AutomationML Server does not need to be configured to match current business processes. Effort is needed to interface to legacy systems but this can be achieved in a significantly shorter time. The methodology proposed by AutomationML Server approach could therefore be considered as a cost efficient solution for realizing the potential in information integration of a company’s production engineering data. Another benefit that is achieved by the suggested approach is that the entire production engineering process can be executed without binding to specific simulation tools. The process is tool-independent and the use of the AutomationML format ensures a transparent execution of the process using any kind of commercial tool. AutomationML is a relative new initiative and thus, it lacks a number of supporting tools that convert different data formats to and from this open format. This practically limits its use to specific data exchange occasions. However, it is expect that due to industry support and interest in this initiative, these data conversion tools will be developed and be supplied by individual stakeholders. Additional research and development effort is required to deal with the issue of data consistency and redundancy, which is not addressed by the current implementation effort. Acknowledgements This work has been partially funded by the EC NMPProject MyCar, Flexible assembly processes for the Car of the Third Millennium. The authors wish to express their gratitude and appreciation to all the project partners for their valuable cooperation and contribution.
S. Makris and K. Alexopoulos / Procedia CIRP 2 (2012) 22 – 27
References [1] Schloegl, W., 2005, Bringing the Digital Factory into Reality – Virtual Manufacturing with Real Automation Data, to appear in CARV 2005 International Conference on Changeable, Agile, Reconfigurable and Virtual Production. [2] Chryssolouris G., Mourtzis D., Papakostas N., 2004, Information Technology in Manufacturing, IFAC Conference on Manufacturing, Modelling, Management and Control., October 21-22 2004, Athens. [3] Hinrichs, J., Rittscher, J., 2004, Collaborative ramp-up planning and controlling, INDIN '04. 2004 2nd IEEE International Conference on Industrial Informatics, 24-26 June 2004, 180 – 183. [4] Bley, H., Franke , C., 2004, Integration of Product Design and Assembly Planning in the Digital Factory, CIRP Annals, 53/1:2530. [5] Chryssolouris, G., 2005, Manufacturing Systems-Theory and Practice. 2nd ed. New York: Springer-Verlag. [6] Schmidgall, G., Kiefer, J., Bär, T., 2005, Objectives of integrated digital production engineering in the automotive industry, 16th IFAC World Congress, Proceedings. [7] Siller, H. R., Estruch, A., Vila, C., Abellan J. V., Romero, F.,2008, Modeling workflow activities for collaborative process planning with product lifecycle management tools, J Intell Manuf,19:689– 700. [8] Seung Hak Kuk , Hyeon Soo Kim, Jai-Kyung Lee, Seungho Han, Seong-Whan Park., 2008, An e-Engineering framework based on service-oriented architecture and agent technologies Computers in Industry, 59:923–935. [9] Flavio Corradini, F., Mariani, L., Merelli, E., 2004, An agentbased approach to tool integration, Int J Softw Tools Technol Transfer, 6:231–244. [10] Mourtzis D., Papakostas N., Makris S., Xantakis V., Chryssolouris G., 2008, Supply chain modeling and control for producing highly customized products, CIRP Annals - Manufacturing Technology, 57/1:451-454. [11] Butala, P., Ueda, K. (Eds.), 2001, Synthesis and adaptation of multiagent communication protocols in the production engineering domain, 3rd International Workshop on Emergent Synthesis (IWES01), Proceedings: 73-82. [12] Alexopoulos, K., Makris, S., Xanthakis, V., Chryssolouris, G., 2011, A Web-services oriented workflow management system for integrated digital production engineering, CIRP Journal of Manufacturing Science and Technology, 4, 290–295. [13] AutomationML, 2012, http://www.automationml.org/, accessed online February 2012. [14] Drath, R., Hundt, L., Lüder, A., Peschke, J., 2008, Seamless Automation Engineering with AutomationML, 14th ICE conference, Lisbon Portugal.
27
