An Agent Enhanced Data Collection Model for RFID Mold Management

Size: px
Start display at page:

Download "An Agent Enhanced Data Collection Model for RFID Mold Management"

Transcription

1 12-ICIT 9-11/4/07 in RoC Going for Gold ~ Quality Tools and Techniques Paper #: Page- 1 /6 An Agent Enhanced Data Collection Model for RFID Mold Management Wei-Ling Wang 1 Dr. Ching-Jen Huang 2 Dr. Chiao-Tzu Huang 3 1,2,3 Department of Industrial Engineering and Management, National Chin-Yi University of Technology, Taichung, Taiwan wlwang@ncut.edu.tw 2 3 cjhuang@cjhuang.idv.tw huang501@ncut.edu.tw ABSTRACT The mold management problem arises in many industries. Examples include the injection manufacturing process and machine tools cutting process in industries. Recently, integrating RFID and agent technology is an emerging, novel approach for effective mold management. In this paper, we propose an agent-enhanced data collection model for RFID mold management system (ARMS). ARMS uses a site monitoring agent group to collect real-time mold data sent from RFID reader and a global manager agent to interact with site monitoring agents and parses mold data for further management decision. ARMS supports the collaborative and autonomous mold data management and moves up the synergy of productive quality. Keywords: RFID, Mold management, Agent 1.0 Introduction The mold management is an important issue for manufacturer to meet customer s needs. Generally, benefits on effective mold management include lower mold inventories, shorter lead times, lower costs, higher productivity, good quality, and greater customer satisfaction. However, in fact, there are just few companies who have mold management. Furthermore, in traditional, all mold data are collected by operator without applying any information technology. For the pass years, radio frequency identification (RFID) technology has been developed, an identification tag containing electronic code, which carry the information about the products that RFID tag attach with. Induced tags by radio wave will transmit these data by radio to a reader. The reader decodes and validates the identification signal prior to transmitting the associated data to a collection system. The advantages of RFID include: (1) identification does not depend on physical contact or direct line of sight observation by the reader, (2) much more data can be contained in the tag than with most automatic data collection (ADC) technologies. RFID technology plays a crucial role in automatic information retrievals. RFID application systems have the potential to provide the real-time data needed to implement enterprise feedback functionality. Recently, the concept of agent has become increasingly important in both artificial intelligence and computer science. An agent is a software entity as an autonomous, goal-directed, computational process capable of robust and flexible interaction with its environment. It can autonomously perform routine tasks with a degree of intelligence to support dynamically changing systems. Now, multi-agent systems are gaining wide acceptance in both industry and academia as a powerful paradigm for coordination in distributed environment. This paper intends to apply RFID and agent technology as a solution for effective and efficient mold management. It presents an agent-enhanced RFID mold management system (ARMS) for implementing real-time mold management. The proposed model uses IT-based data collecting system as a control and integration mechanism to overcome the barriers of distributed data process and timely provides information. The architecture can improve mold management significantly, and move up the synergy of productivity. The remainder of this paper further describes the structure of ARMS architecture. Section 2 reviews the background and previous literature. Section 3 describes the industrial scenario of RFID mold management. Section 4 defines the ARMS architecture in details. The last section concludes our contributions and addresses suggestions for future works.

2 12-ICIT 9-11/4/07 in RoC Going for Gold ~ Quality Tools and Techniques Paper #: Page- 2 /6 2.0 Literature review 2.1 Agent Tang [2004] said it s a new collaborative design environment to facilitate active die-maker involvement in metal stamping product development. Using the agent-based approach, a multi-agent-based system is constructed to integrate die-maker s activities into customer product development process within a distributed, collaborative and concurrent environment. Balasubramanian et al. [1996] proposed a multi-agent approach to concurrent design, manufacturability analysis, process planning, routing and scheduling. Shen et al. [1996] acknowledged that distributed intelligent design environment (DIDE) was a distributed architecture for integrating multidisciplinary engineering tools in an open environment, organized as a population of asynchronous cognitive agents. Sun et al. [2001] proposed a distributed multi-agent environment for product design and manufacturing planning. 2.2 RFID Technology Hodges et al. [2005] assumed that RFID technology is based upon the use of electromagnetic waves, specifically radio frequency electromagnetic emissions, to exchange data between RFID tags and readers. Hodges et al. [2005] and Angeles [2005] stated that readers consist of a transponder and an antenna. Signals are sent from the reader and are responded to by the tags. Again, since the signals sent are in the radio wave frequency, it is not necessary to have line of sight for readers to gather data from the tags. Readers can be either stationary or mobile. RFID is an electronic tagging technology that allows an object, place, or person to be automatically identified at a distance without a direct line-of-sight, using an electromagnetic challenge/response exchange [Want, 2004]. Typical applications include labeling products for rapid checkout at a point-of-sale terminal, inventory tracking, animal tagging, timing marathon runners, secure automobile keys, and access control for secure facilities. Tags can be active, passive or semi-active. Passive tags depend entirely upon the signals sent from the RFID reader for power, much as land-line telephones use power from the telephone signal to maintain functionality. Active tags are independently powered by a battery incorporated in the tag and can autonomously emit signals. Semi-active tags use a battery to supplement the power sent in the reader signal, which improves signal range. 2.3 Mold Management Li et al. [2005] presented a collaborative application portal which provides an application platform for collaborative product design, product data management, manufacturing information management and e-business services for the mold industry. Wong et al. [2002] had concluded the mold industry is an important support industry that comprises primarily of SMEs and can be classified into: (i) tooling and mold making, (ii) part production, such as molding and stamping, and (iii) standard component and tool supply. Tom [1999] had found that the web server can forward or receive the information processed by back-end system using common gateway interface (CGI), active server pages (ASP), Java server pages (JSP), Java servlet and Java script. 3.0 Scenario of RFID Mold Management Mold is a key component for manufacturing in many industries. In traditional, in order to manage mold, all mold data which include mold's basic data, production status, location, maintenance and repair record, and history usage are collected by operator without applying any information technology. Every operator collects mold data before and after production with manual. The process is ineffective and inefficient. The disadvantages are data error by manual, not real time data and not integrated data. Based on the computer development, the mold management takes the advantage of information technology has a new computerized mold management system. Every operator collects mold data before and after production with manual then entries into computer. The operator can inquire and maintain all mold data from the database before and after production. Through the process, the operator knows which mold will be on line, which mold will be maintained, and which mold will be phase out. The disadvantages are still existed. There are data entry errors by manual and not real time data. In this research, we present an agent-enhanced RFID system for implementing real-time mold management. For automatic data collection purpose, we issued a RFID tag on each mold. After

3 12-ICIT 9-11/4/07 in RoC Going for Gold ~ Quality Tools and Techniques Paper #: Page- 3 /6 manufacturing order (M/O) is released, the operator needs to prepare material or part for M/O. Before the mold is set up for matched machine, the operator refers mold history usage data for M/O from computer through wireless and searches mold in mold rack for M/O from RFID tag. When set-up is complete then production is going on. The operator records the mold usage data on history usage data file through wireless. If the mold is not qualified, it will be transferred to R/D for revision. The process is shown in Figure 1. The proposed model uses IT-based data collecting system as a control and integration mechanism. It can overcome the barriers of distributed data process and timely provide information. The advantages are real time data, integrated data and data collected automatically. Manufacturing order (M/O) Released Operator prepares material or part for M/O Operator refers mold history data for M/O from computer through wireless Operator searches mold in mold rack for M/O by RFID tag Operator moves mold and sets up it to matched machine for production Mold is ok Yes No R & D Department makes necessary maintenance Production Operator records the mold usage data on history data file through wireless Figure 1: Process of RFID Mold Management 4.0 The proposed agent-enhanced RFID system 4.1 Architecture of agent-enhanced data collection model for RFID mold management In this paper, we propose an agent-enhanced data collection model for RFID mold management as shown in Figure 2. The system operates based on a two-tier system architecture for data collection and control. The first tier is data collection layer which is focused on receiving mold data grabbed by RFID reader on shop floor. It s an agent-based system by using site monitoring agent (SMA) group as a receiver to catch the RFID reader s data and then processes and stores them to ARMS database. The second layer is application coordination layer which is devoted to backend information management. It uses global manager agent (GMA) to interact with SMA group and parses the data sent from SMA group. These agents work collaboratively and autonomously in real time scenarios. 4.2 RFID process model for ARMS mold management The RFID is an automatic identification technology, relying on real-time remotely retrieving data using devices called active/passive RFID tags or transponders. RFID provides a contactless data accessing

4 12-ICIT 9-11/4/07 in RoC Going for Gold ~ Quality Tools and Techniques Paper #: Page- 4 /6 solution. If a mold is attached with RFID tag, then the mold information will be read by the RFID reader and feedback to the backend information processor. Thus, the manual data input is not necessary and can be eliminated. Taking into consideration of data collecting technique and mold management environment, in order to convey any status changes in real time, the ARMS system use RFID technology to transform the physical processes and statuses into data flow. For the RFID application in this paper, each mold is issued an RFID tag, which contains information about the mold's production status, location, repair record, and history usage. By placing the tag in a special RFID tag reading tray, mold data can be instantly collected. On shop floor level, each mold is arranged to the adjacent area around the right machine for mold management. Therefore, the RFID readers should be placed in distributed locations, which will cause the difficulty of data collection and affect the performance of data process. ARMS applies agent technology to configure dynamic environment flexibly. Figure 2: Architecture of agent-enhanced RFID mold management system (ARMS) 4.3 Agent behavior model in ARMS In this paper, we apply agent technology to develop a timely, autonomous data collection. There are two types of agents in the ARMS system, which operate autonomously and cooperate with each other to accomplish their predefined goals. The two types of agents are defined as follows. The responsibility of SMA is to perform data collection tasks. In order to enable a SMA to create correct messages to reflect incoming RFID tag information, the RFID readers need to be configured with the facility or to be arranged around the manufactory. All readers need to configure with specific type to represent the position in manufactory. During the runtime, after the system is initiated, a SMA will establish connections to all the predefined/adjacent readers in its configuration. When a mold enters a facility, e.g., a machine tools center, the reader will detect and read the tag content information. Then, the middleware will be triggered to collect the data and send message to SMA. The corresponding SMA will receive the tag content and the reader ID from RFID middleware and conveys all changes to ARMS database automatically. The GMA is the core agent of the framework; it manages all of the other agents, control the process logic, as well as supports system operation. When a member of SMA group starts, it should first notify GMA what it is and then send its configuration data to GMA for recognition. GMA then sends instructions to SMA and waits for further mold data coming. Agents need agent communication protocol to interaction with each other. Agent communication model is used to identify the interaction model among agents. In this paper, the model is presented by using the UML sequence diagram. The example of agents interaction model between SMA group and

5 12-ICIT 9-11/4/07 in RoC Going for Gold ~ Quality Tools and Techniques Paper #: Page- 5 /6 GMA are presented in Figure 3. The agents use a common shared communication protocol, agent communication language (ACL) based on the FIPA specifications, among them to understand each other freely and interact with other agents continually. When ARMS starts, each member of SMA group will send a Notify TaskStatus INFORM-type ACL message to GMA to make sure the connection with GMA is O.K. GMA will send back a Get TaskStatus CONFIRM-type ACL message to reply a confirmation. If SMA receives mold RFID data and stores them in ARMS database, SMA will send a Stores RFID data INFORM-type ACL message to GMA. GMA will replies an Agree AGREE-type ACL message to SMA and then start to process the mold transaction data. All of these agents work and react depending on their incoming message, internal state, defined functions and believes. Furthermore, by interaction between SMA group and GMA, structured decision can be made automatically, which are faster and more consistent than human decision making for collaboration effectively and quality. 4.4 System implementation Figure 3: The interaction model between SAM group and GMA The agent system of this paper is based on the Java Agent DEvelopment Framework (JADE) developed by TILab. This software framework follows the agent communication language (ACL) specifications proposed by Foundation for Intelligent Physical Agents (FIPA) and provides a set of graphical tools that supports the debugging and deployment phases. All agents are written in the Java language using the JADE platform. The ARMS system is implemented separately on two JADE container, i.e., manager container residing in the JADE main container, and shop floor container acting as an independent normal container. Manager container with a single agent GMA and shop floor container with SMA group are installed at respective server. Both servers are running the Microsoft Windows 2000 server platform with Java Development Kit 1.5 and TILab JADE version 3.3. The database systems are Microsoft SQL server When the system boots up, agents in manager container and shop floor container will register to JADE main container and DF in main container will record what services the agents support. Agents will cooperate with each other by means of Intranet for security consideration. 5.0 Discussion and Conclusion In order to succeed in the competitive marketplace, enterprises apply IT tools to reflect changing demands and to make new products in a timely matter. For controlling the manufacturing flow effectively and efficiently, many researches depict the significant benefits of using IT-based RFID and agent system as a control and integration mechanism to enable data collection during production process. This paper presents a 2-tier agent-enhanced RFID mold management system for implementing real-time mold data collection. By applying agent technology to establish the information communicating mechanism in distributed environments and using RFID reader as a data collector, the system overcomes the barriers of distributed data process and timely provide information. The advantages are real time data, integrated data and data collected automatically.

6 12-ICIT 9-11/4/07 in RoC Going for Gold ~ Quality Tools and Techniques Paper #: Page- 6 /6 This paper provides the following main contributions: 1. With the aids of agent technology, ARMS presents an agent-enhanced data collection model to solve the problem of distributed data process. 2. By applying RFID technology, ARMS demonstrates an application of RFID technology to mold management. The benefit of mold with RFID tag corresponding to other automatic identification technologies, such as bar codes, is the ability to read a tag without requiring line of sight (depending on the RFID frequency range). 3. For mold management, the system ensures the mold data being accurately automatic updated in real time. References Angeles, R. [2005]. RFID Technologies:, Supply-Chain Applications and Implementation Issues, Information Systems Management, Vol. 22(1), pp Balasubramanian, S., Maturana, F. and Norrie, D.H. [1996]. Multi-agent planning and coordination for distributed concurrent engineering International Journal of Cooperative Information Systems, Vol. 5 (2 3), pp Hodges, S., Harrison, M. [2003]. Demystifying RFID: Principles & Practicalities, Cambridge, United Kingdom: Auto-ID Centre, Institute for Manufacturing, University of Cambridge, Retrieved March 5, 2006, from Li, M.,Wang, J.G.,Wong, Y.S. and Lee, K. S. [2005]. A collaborative application portal for the mould industry Int. J. Production Economics, Vol. 96, pp Shen,W.M., Barthes J.P.A. [1996]. An experimental multi-agent environment for engineering design, International Journal of Cooperative Information Systems, Vol. 5 (2 3), pp Sun, J., Zhang, Y.F. and Nee, A.Y.C. [2001]. A distributed multi-agent environment for product design and manufacturing planning, International Journal of Production Research,Vol. 39 (4), pp Tang, D. [2004]. An agent-based collaborative design system to facilitate active die-maker involvement in stamping part design Computers in Industry, Vol. 54, pp Tom, V. [1999], Enterprise JavaBeans, Addison Wesley Longman Inc., Reading, pp Want, R. [2004]. The Magic of RFID. ACM Queue, Vol. 2(7), pp Wong, Y.S., Wang, J.G. [2002], Domain-specific portal for the manufacturing industry in Singapore needs analysis, Proceedings of the International Manufacturing Leaders Forum Leadership of the Future in Manufacturing, Adelaide, Australia, Feb. 8 10, pp Authors Background Wei-Ling Wang is a lecturer of Industrial Engineering and Management at National Chin-Yi Institute of Technology, Taichung, Taiwan. He was also an instructor of EAN-Taiwan. His research areas of interest include automatic data identification system, evaluation approach, and automatic production system. Dr. Ching-Jen Huang is an associate professor of Industrial Engineering and Management at National Chin-Yi Institute of Technology, Taichung, Taiwan. He received his Ph.D. degree in industrial engineering and engineering management from National Tsing Hua University, Hsinchu, Taiwan. His research interests are in the areas of collaborative design, ebusiness, Information system development, and information management. Dr. Chiao-Tzu Huang is currently an associate professor of Industrial Engineering and Management at National Chin-Yi Institute of Technology, Taichung, Taiwan. He received his Ph.D. degree in industrial department from University of Texas at Arlington, USA. His research interests are statistical process control, automation, and quality management.