Case Study Improving Manufacturing Operations Efficiency Using RFID

On the Web sun.com/rfid

Improving Manufacturing Operations Efficiency Using RFID

A Sun Java System Case Study

Organization Sun Microsystems Newark Manufacturing Facility Key Challenges Can RFID technology be used on the factory floor? Issues include: A large amount of metal used in products and around the production line could potentially block RFID signals. What type of tags should be used, and how should they be placed on inventory and travel carts? Multiple radio sources could interfere with RFID tags. Will RFID tags be read accurately? Solution The six-week pilot showed: 99.5 percent accuracy, with only one missed read No unintentional reads No interference from other radio sources Business Results An internal ROI analysis suggested that a production deployment would provide positive

Sun on Sun: Piloting RFID Technology in Manufacturing Radio frequency identification (RFID) technology is marching inexorably into the supply chain, pushed by mandates from Wal-Mart and the U.S. Department of Defense. Sun is a pioneer in RFID technology, and in December 2003 completed a successful pilot implementation at its Newark, California manufacturing facility. This document describes this effort, including the goals, challenges, and results. Electronic Product Code The Internet of Things RFID tags are an enabling technology in the emerging Electronic Product Code (EPC) network the next generation of bar codes. Promoted by the EPCglobal standards organization in cooperation with the Uniform Code Council, RFID tags are designed to uniquely identify virtually any object, without line-of-sight access or human intervention. When a tagged object is scanned by a reader, the information is processed by the Sun Java System RFID Event Manageran implementation of the EPCglobal Savant standard with enhanced functionality. The Java System RFID Event Manager performs many functions, including data lookup, filtering, and connectivity into back-end IT systems. For example, a reader can determine that a product has moved out of a test area, and forward this information to enterprise application systems for warehouse and supply chain management. A Java System RFID Event Manager can be distributed over multiple compute nodes with no single point of failure; however, a distributed configuration was not used in the pilot. Fully utilized, EPC data can be used to automatically identify objects, monitor location and movement, and trigger events. A near real-time view throughout the supply chain can enable significant gains in operational efficiencies, such as tighter inventory control, increasing throughput, and lowering costs. The Environment EPC networks are an emerging technology, and enterprises are well-advised to undertake proof-of-concept tests or pilot operations while integrating this technology into their production operations. Suns Newark facility manufactures midrange and high-end servers. To maintain the highest levels of quality, these servers undergo a well-designed production process before being shipped to customers. Carts containing server components, such as the chassis, board, and drive, move throughout the factory. At each stage of the production process, workers spend one to two minutes entering data about each system into terminals.

benefits, such as: Reducing transaction times and increasing

line uptime Improving inventory accuracy

Saving operator time when moving products

between manufacturing stations

Case Study Improving Manufacturing Operations Efficiency Using RFID

P2

Its difficult to quantify the cost of a line down due to inaccurate inventory levels or the root cause of shortages. RFID can reduce or eliminate these issues, and improve overall factory efficiency. This will enable us to maintain throughput at lower costs. David Moreira, Manufacturing Engineer, Sun Microsystems

There were also secondary goals, such as determining read accuracy and throughput in a production environment; enterprise information systems (EIS) integration capabilities; and gaining hands-on experience. This pilot was a step toward a larger goal of determining if an EPC network could be implemented throughout a factory, which is expected to eliminate manual key entry and inventory counts while providing item-level tracking and history. The pilot system distinguished between four different components: Two server chassis and two I/O boards. These typically moved into and out of the area in various combinations. The Process The point in the manufacturing flow chosen for the pilot was called the Rotational Capital process. After testing was completed, products were held in a storage area. All rotational material stayed within the Newark facility throughout its testing period. This facilitated RFID testing in a closed-loop environment, without any potential for impacting customer deliveries. The manufacturing process was modified as follows. 1 Components were selected for rotation into the test and tagged with an RFID tag. 2 Material was removed from its stocking location and placed in front of the gate footprint. 3 An operator verified a correct read and accepted the transaction. 4 Items were taken into the test area, and tests were completed. 5 Components were removed from the test bay and positioned in the gate footprint. 6 An operator verified a correct read and accepted the transaction. 7 Components were restocked in their original location.

Periodically, inventory is counted manually to determine exact inventory levels and location. The Goal and Challenges The pilot implementation had one clear goal: Could RFID technology be deployed successfully on the Newark factory floor? There were many issues to address: Metal: Both the products and the carts used to transport them around the factory floor were made primarily of metal, which blocks radio waves. In addition, directly attaching a tag to metal could create difficulties in reading it. Tags were placed on system components located on an interior surface of the metal chassis, providing further reading difficulties. Orientation and placement: Tag orientation (with respect to the reader antenna) and placement on the product had an affect on tag-reading ability. Interference: The factory floor was filled with radio waves from other sources, including cellular phones and 802.11 wireless devices. The pilot verified that these would not interfere with EPC network tags. Stray or missed reads: The pilot was intended to prove that the readers could sense all tags that were intended to be read, would not miss any reads, and would not unintentionally pick up stray reads from units that were passing by.

Case Study Improving Manufacturing Operations Efficiency Using RFID

P3

Hardware The hardware components primarily consisted of three components: Gate, reader, and server. Gate: All traffic into and out of the Rotational Capital process areas passed through one point. A gate, or fixture which mounts the reader and antenna, was constructed at this location. To limit the amount of metal in the reader area, the gate was constructed of 3.5inch and 2.5-inch PVC plastic. A single Alien Technology 915-MHz reader with two linear polarized antennas was attached. Dual antennas facilitated tag reading for both incoming and outgoing carts. Tags used in the proof of concept were passive Alien Technology I-tags. They exhibited the best performance and have a form factor that could be easily accommodated in the components chosen for the proof of concept. I-tags could be vertically oriented on both the components and cart, which allowed the use of linear antenna. In the prototype testing, this exhibited much better performance than typical circular tags and antennas. Tags were mounted on system chassis and major components (such as I/O boards). A Sun LX50 server ran the software and displayed read results to the operator. A Sun Enterprise 450 server was used to run the Java System RFID Information Server and unit tracking application.

Software Software used in the pilot was built on the Sun Java RFID System Suns RFID software stack and included the Java System RFID Event Manager (Savant) and Java System RFID Information Server. A custom unit tracking application was developed that included two components: A user interface front end and a business logic back end. Both the Java System RFID Information Server and the unit tracking application were built based on the Java System Application Server. The overall system maintained track and trace capabilities for location history, and provided a real-time inventory within the Rotational Capital process area.

The Java System RFID Event Manager provided an interface to the reader, filters, and process RFID reads, and served as a generalpurpose routing system. It helped integrate RFID event data with an EIS by defining a set of interfaces that facilitated sending and receiving data in real time. Java System RFID Event Managers were meant to be part of a distributed architecture with no single point of failure; however, only a single Java System RFID Event Manager was used in this pilot. Data from the Java System RFID Event Manager fed into the Java System RFID Information Server, where it was stored and made available in a consistent manner to any

Figure 2: Suns RFID Software Stack

application that needed it. Data was stored persistently on an Oracle relational database and exposed through an HTTP interface. The unit tracking application was built to enable nontechnical operators to interact with the RFID system, accessing information such as registration, confirmation, and incoming or outgoing direction. The application was designed to be intuitive, with easy-to-read pictures and tables for each transaction. Operators could

The pilot also relied on software from third parties, including an Oracle relational database and a reporting module from Brio. The Brio tool enabled users to visualize raw data that was gathered during the pilot and make business sense out of it by using built-in query and statistical analysis functions.

Figure 1: I-tag Placement on a PCI Board

Case Study Improving Manufacturing Operations Efficiency Using RFID

P4

The Java System RFID Event Manager and RFID Information Server work with industry-standard readers, processing events and making them available to existing EIS applications from Oracle, SAP, and applications from other leading vendors. This comprehensive solution is ideal for companies that want to integrate RFID information into their production applications. Julie Sarbacker, Director, RFID Business Unit, Sun Microsystems

verify stock numbers, quantities, specific incoming or outgoing locations, and add or deselect transaction units. The application had three user modes: Demo, manager, and operator. User type was recorded as part of each transaction.

Especially surprising was the read rate of I-tags located on I/O boats. These were expected to be the most difficult to read because they were placed in an indentation surrounded by metal and perpendicular to the antennas.

Figure 3: Application User Interface

Determining an ROI was not a stated goal of the pilot, but there were many indications that an RFID environment would not only be productive, but also financially beneficial. Automatically inputting data at each process station could save as much as 90 seconds of operator time, eliminate many if not all inventory counts, and reduce or eliminate root causes of inventory variances. Internal ROI

The Results The pilot ran for six weeks, and generated hundreds of RFID transactions. With a read accuracy of 99.5 percent (just one missed read), the goal was met. There were no instances of unintended transactions or failing to sense tags that were positioned under the gate. This was especially encouraging given the high degree of metal present in the products.

Case Study Improving Manufacturing Operations Efficiency Using RFID

P5

The Java System RFID Event Manager is built on a federated services architecture, with no single point of failure. This means if a forklift accidentally takes out an Event Server, its work is picked up by other EPC servers on the network. Sumith Mathew, Staff Engineer, Sun Microsystems

calculations pointed to a distinct opportunity for reducing the effort required to keep track of material throughout the factory at a reasonable cost. While application performance was not part of the pilot objectives, response time was acceptable for an interactive application. Results from the pilot indicated that the software could easily accommodate a full production load of tens of thousands of reads per quarter. Lessons Learned The goals of the pilot program were not only to determine if RFID technology could be used in a production environment, but also to gain an understanding of principal issues. The pilot helped both the manufacturing engineering and software development teams in their RFID efforts. Conclusions included: Tag placement was critical, and required empirical testing to find the optimal location. Initial testing was done with omni-directional tags and antennas, but lab testing showed they were not as accurate in this environment. Properly placed I-tags offered highly accurate reads when used with linear antennas. In addition, tags had to be isolated from metal components. Until tags are manufactured with an insulating material, there will be restrictions about where they can be placed. The pilot highlighted the necessity of training factory floor operators. Despite best efforts to develop an easy-to-use application that saves time when moving between stations, there were failures due to operators neglect-

ing to transact items as indicated in the proof-of-concept instructions. Tags were too big for some of the valuable components used in production, such as memory DIMMs and network interface cards. The continued reduction of tag dimensions will enable these components to become more widely used in the manufacturing process. While RFID-enabled, serial-level data collection gave greater visibility into manufacturing processes, not all of this data could be used due to existing constraints. For example, existing EIS applications may not be able to accept this data, and processing it may require more IT resources. During the planning phase, it will be necessary to understand how much of this data is useful and how it can be integrated with existing EIS applications. Standards organizations have defined EPC tag data length and format. Choosing the correct standard was important because there were existing part and serial numberingschemes. Some businesses may have proprietary numbering schemes in place, while others may use more standard Global Trade Item Number (GTIN) formats. It is likely that EPC numbers will need to coexist with the existing ones. Choose a tag data standard that offers the data length and format that can accommodate your scheme. The cost of the RFID gate, including reader and antennas, was very reasonable. For this pilot, material costs were under $5000. This compared favorably to handheld barcode scanners that typically cost $2000 each.

Case Study Improving Manufacturing Operations Efficiency Using RFID

On the Web sun.com/rfid

Improving Manufacturing Operations Efficiency Using RFID

Next Steps The next step will be to evaluate the feasibility of a production deployment of an RFID system on the factory floor, with a goal of minimizing periodic inventory counts. While this would require integration of the application with the Oracle manufacturing application, much of the work has already been done, so it is expected to be a minimal effort. The successful conclusion of the pilot program also suggested new opportunities, for example, an internal pilot for high-volume transactions. Additional benefits could be obtained by piloting a process with a higher transaction volume for increased efficiency. The next process to be piloted could be the replenishment of material in the assembly areas from central stocking locations. Material for this process naturally moves through choking points, where it would be easy to install the required RFID antennas and reader. Another opportunity would be to undertake a pilot with a Sun supplier. There is a potential for great benefits by having the material tracked across the supply chain. It could greatly streamline receiving and stocking processes as well as reduce the time needed for this material to become available in assembly areas. Get the details. For more information on Suns RFID solutions, please visit sun.com/rfid. Pilot Components Alien Technology 915-MHz Reader Two Linear Polarized Antennas Sun Enterprise 450 Server Sun LX50 Server Solaris 9 Operating System Java System RFID Event Manager Java System RFID Information Server Oracle RDBMS Brio Reporting Module Custom RFID User Applicationr

Sun Microsystems, Inc. 4150 Network Circle, Santa Clara, CA 95054 USA Phone 1-650-960-1300 or 1-800-555-9SUN Web sun.com

Sun Worldwide Sales Offices: Africa (North, West and Central) +33-13-067-4680, Argentina +5411-4317-5600, Australia +61-2-9844-5000, Austria +43-1-60563-0, Belgium +32-2-704-8000, Brazil +55-11-5187-2100, Canada +905-4776745, Chile +56-2-3724500, Colombia +571-629-2323, Commonwealth of Independent States +7-502-935-8411, Czech Republic +420-2-3300-9311, Denmark +45 4556 5000, Egypt +202-570-9442, Estonia +372-6-308-900, Finland +358-9-525-561, France +33-134-03-00-00, Germany +49-89-46008-0, Greece +30-1-618-8111, Hungary +36-1-489-8900, Iceland +354-563-3010, IndiaBangalore +91-80-2298989/2295454; New Delhi +91-11-6106000; Mumbai +9122-697-8111, Ireland +353-1-8055-666, Israel +972-9-9710500, Italy +39-02-641511, Japan +81-3-5717-5000, Kazakhstan +7-3272-466774, Korea +822-2193-5114, Latvia +371-750-3700, Lithuania +370-729-8468, Luxembourg +35249 11 33 1, Malaysia +603-21161888, Mexico +52-5-258-6100, The Netherlands +00-31-33-45-15-000, New ZealandAuckland +64-9-976-6800; Wellington +64-4-462-0780, Norway +47 23 36 96 00, Peoples Republic of China Beijing +86-10-6803-5588; Chengdu +86-28-619-9333; Guangzhou +86-20-8755-5900; Shanghai +86-21-6466-1228; Hong Kong +852-2202-6688, Poland +48-22-8747800, Portugal +351-21-4134000, Russia +7-502-935-8411, Singapore +65-6438-1888, Slovak Republic +421-2-4342-94-85, South Africa +27 11 256-6300, Spain +34-91-596-9900, Sweden +46-8-631-10-00, SwitzerlandGerman 41-1-908-90-00; French 41-22-999-0444, Taiwan +886-2-8732-9933, Thailand +662-344-6888, Turkey +90-212-335-22-00, United Arab Emirates +9714-3366333, United Kingdom +44-1-276-20444, United States +1-800-555-9SUN or +1-650-960-1300, Venezuela +58-2-905-3800 2004 Sun Microsystems, Inc. All rights reserved. Sun, Sun Microsystems, the Sun logo, Java, the Java Coffee Cup logo, Solaris, and Sun Enterprise are trademarks or registered trademarks of Sun Microsystems, Inc. in the United States and other countries. Information subject to change without notice. 03/04 R1.0

SUN