Citation
GatorPacker(TM): A Worker Productivity Monitoring System Using RFID (Radio Frequency Identification) Technology

Material Information

Title:
GatorPacker(TM): A Worker Productivity Monitoring System Using RFID (Radio Frequency Identification) Technology
Creator:
THYAGARAJA, KARTHIK ( Author, Primary )
Copyright Date:
2008

Subjects

Subjects / Keywords:
Antennas ( jstor )
Bar codes ( jstor )
Boxes ( jstor )
Conveyors ( jstor )
Databases ( jstor )
Employee productivity ( jstor )
Productivity ( jstor )
Radio frequency identification ( jstor )
Sensors ( jstor )
Software ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Karthik Thyagaraja. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
5/31/2010
Resource Identifier:
659898614 ( OCLC )

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

GATORPACKERTM: A WORKER PRODUCTIVITY MONITORING SYSTEM USING RFID (RADIO FREQUENC Y IDENTIFICATION) TECHNOLOGY By KARTHIK THYAGARAJA A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ENGINEERING UNIVERSITY OF FLORIDA 2007

PAGE 2

Copyright 2007 Karthik Thyagaraja

PAGE 3

To Dr.We lt, Kiran, Kavitha and my parents.

PAGE 4

iv ACKNOWLEDGMENTS I would like to thank my advisor, Dr. Bruce Welt, for introducing me to the exciting field of radio frequency identific ation (RFID). The insights and knowledge I have gained since then in the area and in implementing GatorPackerTM system would not have been possible but for Dr. WeltÂ’s guida nce and help. I would like to thank Dr. Howard Beck and Dr. William R. Eisenstadt for their support and help and for serving on my committee. I would like to thank Mr. Juan Vila a nd Mr. Brian Bishop for helping me in conducting the pilot tests. Thanks are al so due to Ms. Mary Hall and Mr. Billy Duckworth, who were very helpful to me at different times during my master's research work. I would also like to thank Tyco/Senso rmatic for donating RFID tags, reader and reader antennas that were used during this project; Intelleto Technol ogies for its support in developing LinemasterTM in accordance with specifications required for GatorPackerTM; Datamax Corporation for providi ng I-4210 RFID printer; Seagull Scientific for providing BarTender label printi ng software; and Tri-Tronics Company Inc. for providing sensors. Finally, I would like to thank my parents, my brother Kiran and my sister Kavitha for all their love and support.

PAGE 5

v TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLE..............................................................................................................vii LIST OF FIGURES.........................................................................................................viii LIST OF ACRONYMS.......................................................................................................x ABSTRACT......................................................................................................................x ii CHAPTER 1 INTRODUCTION........................................................................................................1 2 RFID PRIMER.............................................................................................................3 History........................................................................................................................ ..3 RFID System................................................................................................................4 EPCglobal and EPC (Electronic Product Code)...........................................................5 RFID Applications........................................................................................................5 3 MATERIALS AND METHODS.................................................................................7 Hardware....................................................................................................................... 7 LineMasterTM.........................................................................................................7 Barcode Reader.....................................................................................................7 RFID Tags.............................................................................................................8 RFID Printer......................................................................................................... 8 RFID Antenna.......................................................................................................9 Trigger...................................................................................................................9 Serial to TCP/IP Radio..........................................................................................9 Software....................................................................................................................... .9 Tasks Menu.........................................................................................................11 Reports Menu......................................................................................................11 Web Admin Console...........................................................................................12 Special Tools..............................................................................................................12 Packer ID Sticks..................................................................................................12 RFID EPC Calculator..........................................................................................12

PAGE 6

vi Working of the System...............................................................................................13 4 RESULTS AND DISCUSSION.................................................................................26 5 FUTURE WORK........................................................................................................35 LIST OF REFERENCES...................................................................................................36 BIOGRAPHICAL SKETCH.............................................................................................37

PAGE 7

vii TABLE Table page 3-1 SGTIN-96 Tag Format.............................................................................................25

PAGE 8

viii LIST OF FIGURES Figure page 2-1 Passive RFID Tag Architecture..................................................................................6 2-2 96-bit EPC Layout......................................................................................................6 3-1 LinemasterTM............................................................................................................14 3-2 Microscan Model MS-850 Barcode Reader.............................................................14 3-3 Datamax I-4210 RFID Printer..................................................................................15 3-4 Tyco Sensormatic ® Omniwave Antenna.................................................................15 3-5 IDEC’s SA1C-DD3 Sensor......................................................................................16 3-6 Data Hunter’s Serialan Du al RS232 Serial Ports WLAN........................................16 3-7 System Architecture.................................................................................................17 3-8 MDI Parent Window................................................................................................18 3-9 Maintain Menu.........................................................................................................18 3-10 Manage Packers Form..............................................................................................19 3-11 Tasks Menu..............................................................................................................19 3-12 Collect Data Form....................................................................................................20 3-13 Productivity Summary Report Creation Form.........................................................20 3-14 Productivity Summary Report..................................................................................21 3-15.Productivity Summary Report...................................................................................22 3-16 Detailed Productivity Report....................................................................................23 3-17 Packer ID Sticks.......................................................................................................24 3-18 RFID EPC Calculator...............................................................................................25

PAGE 9

ix 4-1 Rising Edge of Signal...............................................................................................30 4-2 Total Number of Boxes Scanned Per Day Using the Trigger..................................30 4-3 Total Number of Boxes Scanned Per Day Without Using the Trigger....................31 4-4 Boxes Very Close to Each Other..............................................................................31 4-5 Changing the Speed of Conveyor Belts to Separate the Boxes................................32 4-6 Condition of Some Packer Tags After Two Months of Use....................................32 4-7 Overall Productivity Data of Each Packer between June 15 and August 1.............33 4-8 Total Number of Scans per Day...............................................................................33 4-9 Cash Flow Diagram..................................................................................................34 4-10 IRR Sensitivity Analysis..........................................................................................34

PAGE 10

x LIST OF ACRONYMS ASP.NET Active Server Pages .NET BNC Bayonet Neill Concelman EAN European Article Number EPC Electronic Product Code GIAI Global Individual Asset Identifier GRAI Global Returnable Asset Identifier GUI Graphical User Interface ID Identifier IRR Internal Rate of Return IFF Identify: Friend or Foe system MDI Multiple Document Interface RFID Radio Frequency Identification ROI Return of Investment RS-232 Recommended Standard 232 SGLN Serialized Global Location Number SGTIN-96 Serialized Global Trade Item Number-96 bit SQL Structured Query Language SSCC Serial Shipping Container Code TCP/IP Transmission Control Protocol/Internet Protocol UCC Uniform Code Council

PAGE 11

xi UHF Ultra-High Frequency UI User Interface WLAN Wireless Local Area Network

PAGE 12

xii Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Engineering GATORPACKERTM: A WORKER PRODUCTIVITY MONITORING SYSTEM USING RFID (RADIO FREQUENC Y IDENTIFICATION) TECHNOLOGY By Karthik Thyagaraja May 2007 Chair: Bruce A. Welt Major: Agricultur al and Biological Engineering This thesis discusses the design and implementation of GatorPackerTM, a worker productivity monitoring system which uses RFID (Radio Frequency Identification) Technology. A major produce packaging company felt the ne ed to monitor the productivity of its workers. The company employs many migran t laborers. The workers are paid minimum wages. Packing vegetables is a low-skill, low-pay job. Low wages tend to motivate workers to be less productive in order to maximize paid work hours. Productivity monitoring provides an opportunity to identif y highly productive workers. The company can minimize costs by using fewer highly productive workers while paying higher base wages plus incentives. A need for a sy stem that could automatically track worker productivity was identified. For this operati on, association of packer and product type produced is crucial for measuring productivit y. Automating data colle ction was desirable. Among different Auto-ID (automatic identifi cation) technologies (barcodes, RFID,

PAGE 13

xiii etc.), RFID technology was viewed as wellsuited because it does not require line-ofsight, which makes it forgiving in terms of how individual workers place identifying tags on work product. Radio frequency identificati on technology virtually eliminated the need for worker training and maximized data co llection accuracy. Therefore, GatorPackerTM was developed to automatically collect data to help companies monitor worker productivity. As part of the thesis, the complete system, involving many hardware and software components, was designed and implemented. Also, the system was deployed in a produce packaging company and pilot tests were conducted. Many deployment issues were solved. The economic analysis of the data collected has shown that, the company where the pilot was conducted, can save about $29,000 per year per production line if all workers work at the rate of the average of the top 5 productive workers, which translates into a payback period of about 1 year and a compelling internal rate of return of 93%.

PAGE 14

1 CHAPTER 1 INTRODUCTION “ There Ain't No Such Thing As A Free Lunch ” [1]. Monitoring productivity of workers and providing incentives based on performance are beneficial to both the employers and productive workers. Systems that efficiently monitor productivity allow companie s to better manage labor costs. Productivity monitoring has b een an issue especially in the produce packaging industry because of the type of labor i nvolved. Migrant laborers are paid minimum wages. A major produce packaging company fe lt the need to measure the productivity of the company's packers. It employs a lot of migrant laborers for packing vegetables. Packing vegetables is a low skill, low pay j ob. Low wages tend to motivate workers to be less productive in order to maximize paid wo rk hours. Productivity monitoring provides an opportunity to identify highly productive workers. Companies can minimize costs by employing fewer, but highly productive worker s while paying higher base wages and/or incentives. A Florida-based vegetable repacker wanted to establish an incentive program to reward its most productive workers, but did not have a means of knowing how many boxes of fruit or vegetables each worker was packing per day. This thesis describes the development and implementation of a worker productivity monitoring system. Currently productivity of each packer is not measured directly. Terminating employees due to low productivity has been co stly due to accusations of discrimination and unfairness. An automated worker productiv ity system may provide the extra benefit of documentation for employee termination purposes.

PAGE 15

2 An initial attempt at automating the data collection using pres s-on barcode labels proved unsuccessful, because labels would stic k to gloves and label backing made a mess in the production line. To overcome the prob lems of press-on labels, RFID (Radio Frequency Identification) was used. RFID does not require line of sight and it can perform better than barcodes under harsh conditions. Therefore the objective of this work was to design and implement such a worker productivity monitoring system using RFID t echnology and to test its effectiveness. The system is expected to help to increase ove rall productivity and help the exceptional workers to be properly rewarded. This syst em should enable more productive workers to earn higher wages.

PAGE 16

3 CHAPTER 2 RFID PRIMER Radio Frequency Identification (RFID) is generally used to describe any technology that uses radio signals to identify specific objects [2]. It is a form of autoidentification (Aut o-ID) technology. History RFID Technology was first used by the Br itish air force durin g World War II to identify its own aircraft from that of the en emies. It was called the IFF or Identify: Friend or Foe system. In the 1970s, the Los Alamos National Laboratory began using RFID to tag and monitor nuclear and other hazardous materials. Los Alamos also developed a passive RFID tag, which was used to track cattle [3]. In the early 1990s, IBM engineers develope d and patented an ultra-high frequency (UHF) RFID system. UHF offered longer read ranges (up to 20 feet under good conditions) and faster data transfer. UHF RFID got a boos t in 1999, when the Uniform Code Council (UCC), European Article Number (EAN) International, Procter & Gamble and Gillette funded establishment of the Auto-ID Center at th e Massachusetts Institute of Technology. The technology that was developed in this laboratory was licensed to the Uniform Code Council in 2003, and the Uniform Code Council created EPCglobal, as a joint venture with EAN International, to commercialize Electronic Product Code (EPC) technology [3]. Now RFID technology has enor mous backing from some of the worldÂ’s largest retailers including Wal-Mart, Target, Tesco, and Metro as well as from the U.S Department of Defense, because of its poten tial for enhancing supply chain visibility.

PAGE 17

4 RFID System An RFID system consists of four elem ents: RFID tags, RFID readers, reader antennas, and information processing software [2]. Software is critical for filtering and analyzing data as well as integrating RFID technology into backend databases. An RFID tag is made up of an antenna, a silicon chip and a s ubstrate (encapsulation material) (Figure 2-1) [4]. RFID tags are cl assified as either passive, semi passive, or active. Passive tags do not have their own source of energy and use the method known as “energy harvesting” to absorb energy from a reader and use it to transmit the data back to the reader. Active tags have their own power s ource but can be read only when they come in the read range of the reader. Semi-passive tags use an internal power source to monitor environmental conditions, but require Radio Fr equency (RF) energy transferred from the reader similar to passive tags to power a ta g response. They are used in situations like cold chains where it is important to track the movement as well as the condition of an item [4]. RFID Readers (or Interrogators) generate signals that provide power for passive tags as well as create an interrogation signal. A tag captures the energy it receives from a reader to supply its own power and then executes commands sent by the reader. The simplest command results in the tag sending back a signal containing a unique digital identification number. Upon r eceiving information from a tag, the reader decodes the information in its decoding software and then transmits it to the information management system. Tag read range is influenced by radio frequency, reader power output, working environment, and tag antenna geometry. Read rates are affected by tag type, dielectric

PAGE 18

5 properties of products, placement, orientation an d reader settings. Sources of interference include cell phones, walkie-tal kies, metals, liquids, etc. EPCglobal and EPC (Electronic Product Code) EPCglobal is a joint venture between EAN International and the Uniform Code Council (UCC). It is a not-for-profit organizat ion entrusted by industr y to establish and support the Electronic Product Code (EPC) Network as the global standard for immediate, automatic, and accura te identification of any item in the supply chain of any company, in any industry, anywhere in the worl d. Its objective is to drive global adoption of the EPCglobal Network. The EPCglobal Network was develope d by the Auto-ID Center at the Massachusetts Institute of Technology (M.I.T.). The EPC Network is a set of services that leverage the existing Internet infrastruc ture for management and application of RFID product data [2]. The electronic product code (EPC) is a new product numbering standard. The EPC is capable of assigning a unique number to any item (e.g., a can of soda could have its own unique EPC number). The EPC allows comp anies to track products at the individual item, case, pallet and/or container level. Th e structure of EPC is shown in Figure 2-2. RFID Applications RFID can be used in any sector of indust ry where data collecti on is involved. Some of the areas where RFID Tec hnology is currently used are Retail supply chain management. Pharmaceutical industry Library and media management applications Asset management

PAGE 19

6 Access control Vehicle parking monitoring Person identification Apparel industry Animal identification/animal tagging Tracking worker productivity (as described in this document) Figure 2-1. Passive RFID Tag Architecture. Figure 2-2. 96-bit EPC Layout Substrate Antenna IC Chip

PAGE 20

7 CHAPTER 3 MATERIALS AND METHODS This chapter describes the hardware setup and implementation details of the GatorPackerTM system. The components that make up GatorPackerTM can be broadly classified into the following categories: Hardware Software Special Tools Hardware LineMasterTM This prototype RFID reader (LineMasterTM) was developed in collaboration with Intelletto Technologies Inc. (Ontario, Canada). It is a dual-functiona l RFID reader that also accepts additional input via RS-232 port (e.g., barc ode scanner). It permits transparency between RFID and barcode and allows the barcode and RFID tag to coexist. Linemaster receives RFID data from the RFID antenna and barcode data from a barcode reader. The Linemaster prototype developed for this work is shown in Figure 3-1. Barcode Reader Microscan Model MS-850 was used to read the barcodes affixed on the boxes (Figure 3-2). These barcodes identify the type of the products. The serial output from this reader is fed into one of the inputs of the LineMasterTM. The advantages of this reader are

PAGE 21

8 its high scan speed, long read range and wide sweep angle which makes it ideal for our application. RFID Tags Ultra high frequency (UHF) 96-bit Cla ss 1 EPC RFID tags were used for identifying each packer (Tyco International (U S) Inc., Princeton, New Jersey). Serialized Global Trade Item Number (SGTIN-96) EPC format was used to encode the packer number. SGTIN is based on Global Trade Item Number (GTIN) code. GTIN identifies a particular class of objects. By adding serial number, GTIN can be used to identify individual objects uniquely. SGTIN is formed by combining GTIN and a unique serial number [5]. SGTIN-96 data stru cture is shown in Table 3-1. The header identifies the type of EPC stru cture (type, version, length). Filter value indicates the type of object (item, case a nd pallet). Partition indicates where the subsequent company prefix and item refere nce numbers are divided. The value of the header section for SGTIN-96 is 48 [5]. The filter section was assigned a value of 3 and the company prefix was assigned an arbi trary value of 999999. The packer number/ID was represented using the item reference section. The serial number was used to represent serial number of a packer stick assigned to each packer. Packer number 3, having serial number 5 is represented in SGTIN-96 format as 48; 3; 999999; 3; 5. The corresponding bit string is 00110000; 011; 110; 11110100001000111111; 000000000000000000000011; 00000000000000000000000000000000000101. The hexadecimal value of this string is 307BD08FC00000C000000005. RFID Printer Labels with RFID inlays were enco ded and printed using DMX I-4210 Datamax Printer (Datamax Corporation, Orlando, Flor ida) (Figure 3-3) th rough BarTender Label

PAGE 22

9 and RFID software (RFID Enterprise Ed ition, Version 7.74, Seagu ll Scientific Inc, Bellevue, WA). RFID Antenna Tyco Sensormatic ® Omniwave Antenna (Sensormatic Electronics Corporation, Tyco ADT, USA), which is an EPC Class 1 Circ ular antenna was used to read RFID tags (Figure 3-4). This antenna wa s connected to the LinemasterTM antenna input after modifying its cable with a female BNC connector. Trigger IDEC’s SA1C-DD3 (IDEC Corporation, US A) photoelectric sensor was used for triggering the Microscan barcode reader. A Tri-Tronics refl ective sensor (Tri-Tronics Company Inc., Tampa, Florida) triggered the LinemasterTM. Both were synchronized to timeout at the same time. Serial to TCP/IP Radio An RS-232 to wireless TCP/IP radio (Seria lan, Data Hunter, California) is used to convert the serial output from the LinemasterTM into TCP/IP (Figure 3-6). The software communicates to this radio through TCP connecti on. The advantage of the radio is that it obviates the need for long serial cables and/or signal boosters. It also removes the data collection workstation from the production ar ea. Finally the TCP/IP communication is more reliable than serial communication. Software The software part can be divided into two components (Figure 3-7): A data acquisition desktop component that captures the RFID data and adds it into the database 96-bit Tyco UHF EPC Class 1 RF ID tags are used for identifying each packer.

PAGE 23

10 A web-based Admin Console that helps the management to keep track of productivity by providing real time informa tion about the workers in the form of raw data, Graphs, Charts, Reports etc. The software was entirely built on the .NET platform [6]. The choice of .NET for implementing the system was due to the following reasons: Ease and speed of development. Easy interaction with Component Object Model (COM) technology. Easy application deploy ment and maintenance. SQL Server Express 2005 database (Micro soft Corp., Redmond, WA) is used for storing the data. Data coll ection is implemented using VB .NET programming language and the web-based Admin console is impl emented using ASP.NET (Microsoft Corp., Redmond, WA). Functionality and User Interface Features The graphical user-interface (GUI) for th e data collection component was designed in accordance with common windows applicat ion design standards (Figure 3-8). The GUI is designed as a Multiple Document Interf ace (MDI) application, wh ich enables users to work with multiple documents at the same ti me. During the pilot, the supervisors were initially not willing to learn new software. But after seeing the familiar MDI interface they were able to learn the software quickly. Core functionalities are organized into three groups, which correspond to the following three Menu Items in the GUI: Maintain Menu

PAGE 24

11 The Maintain Menu (Figure 3-9) provides functionality to add, delete or edit information on workers (packers), products (items) and producti on lines (scanners). Management of these data used the co mmon form design as shown in Figure 3-10. Tasks Menu The Tasks menu (Figure 3-11) provides func tionality to print packer RFID labels and item barcode labels. Each entry in this form corresponds to a production line where the reader is installed. The Collect Data form (Figure 3-12) is used to connect to different radios/scanners. The user can select any numbe r of readers to connect to. Data collected from each radio is added into the central database. Reports Menu Three reports were initially develope d for supervisors including Productivity Summary report, Detailed Productivity Re port and Tag Management report. Figure 3-13 shows the form used to genera te the productivity summary report. It has date and time fields to specify the date range for the report. The two radio buttons specify the type of summary report needed. Th e “By No. of Scans” radio button generates Unit Productivity report and the “By Wei ght” button generates Weight Productivity report. The Unit productivity report, shown in Figure 3-14, gives information on number of boxes scanned by each packer per hour(num ber of scans/hour) whereas the Weight Productivity report, shown in Figure 3-15, gi ves information on the total pounds packed by each packer per hour(total pounds/hour) for the time range specified by the user. The Detailed Productivity report (Figure 3-16) gives detailed information about each packer’s productivity. It gives for each packer ; number and the total weight of each type of item packed, unit productivit y, weight productivity, time of first and last scans for a given day and duration of work.

PAGE 25

12 The Tag Management report is used to ma nage RFID tags. This report can be used to discover vital statistic s about individual tags such as when the tag was created, the number of times it was read within a given time frame and the time when it was last seen. This information is useful in id entifying poorly performing or dead tags. Web Admin Console Currently the Web Admin Console s upports only report generation. In the future functionalities to display real-time metrics may be added. Special Tools Packer ID Sticks Press on RFID labels were designed to s how Packer ID, Stick Serial Number and Date created. These labels were attached to polyethelyne sticks. Th e dimension of each stick is around 18Â’Â’x 1Â’Â’. Pl astic sticks are durable and can withstand harsh and damp conditions. Each stick is unique ly identified by packer numbe r and stick serial number. This information is encoded into the RFID tag as well as printed on the label (Figure 317). RFID EPC Calculator The EPC calculator was developed and used to encode and decode different EPC tag formats like SGTIN, SSCC, SGLN, GRAI and GIAI. The RFID data from the readers is read in hexadecimal format. It is difficult to figure out the type of EPC code by just looking at this hexadecimal value. The EPC cal culator helps in decoding this information. The encode functionality in the calculator can be used to encode desired base 10 EPC encoding into the corresponding hexadecimal va lue. The RFID EPC calculator interface is shown in Figure 3-18.

PAGE 26

13 Working of the System Each packer was assigned about 15 reusable packer ID sticks (passive UHF EPC Class 1 RFID tags). As workers pack boxes, he or she slips a stick into the box. As the box travels down the conveyor, it simultaneously triggers the Linemaster and the barcode reader. The Linemaster seeks the RFID ta g data from the RFID antenna and Item Barcode data from the barcode scanner. The Linemaster was programmed to combine these data into a packet before sending it to the serial output. The pack et is usually in the form “” where the prefix B in the first string indicates barcode and prefix R indicates RFID. If no data are read within the given time, the system times out and sends a packet in the form “”. The serial output from Linemaster is converted to TCP/IP by the RS-232 to wireless TCP/IP radio. These data are transmitted to an 802.11x access point and received by GatorPacker software. The software collects and stores data in a database. Data reports tabulate worker productivity. The database is update d in real time. GatorPacker is a multithreaded application in which data collecti on and report generation run independently and concurrently. Once data are placed into the database the supervisor or management can view reports to find out how each packer is performing. The system should help companies to better manage their workforces. Management can easily see individu al productivity data. The system can also help to alert the supe rvisors about any problems the workers may be facing.

PAGE 27

14 Figure 3-1. LinemasterTM Figure 3-2. Microscan Mode l MS-850 Barcode Reader

PAGE 28

15 Figure 3-3. Datamax I-4210 RFID Printer. Figure 3-4. Tyco Sensormatic ® Omniwave Antenna

PAGE 29

16 Figure 3-5. IDECÂ’s SA1C-DD3 Sensor. Figure 3-6. Data HunterÂ’s Serialan Dual RS232 Serial Ports WLAN.

PAGE 30

17 Database Workstation Business Logic Radio Serial cable Web Admin Console Linemaster Figure 3-7. System Architecture

PAGE 31

18 Figure 3-8. MDI Parent Window Figure 3-9. Maintain Menu

PAGE 32

19 Figure 3-10. Manage Packers Form Figure 3-11. Tasks Menu

PAGE 33

20 Figure 3-12. Collect Data Form Figure 3-13. Productivity Summ ary Report Creation Form

PAGE 34

21 Figure 3-14.Productivity Summary Report by Scans

PAGE 35

22 Figure 3-15.Productivity Summary Report by Weight

PAGE 36

23 Figure 3-16. Detailed Productivity Report.

PAGE 37

24 Figure 3-17. Packer ID Sticks

PAGE 38

25 Figure 3-18. RFID EPC Calculator Table 3-1. SGTIN-96 Tag Format. Section Header Filter Value Partition Company Prefix Item Reference Serial Number No. of Bits 8 3 3 20-40 24-4 38

PAGE 39

26 CHAPTER 4 RESULTS AND DISCUSSION The system was deployed for the pilot phase on May 25, 2006. Initially, no trigger was used to detect a box when it cam e into the range of the LinemasterTM. The problem with this was that the system could not differentiate between products (e.g., 4# box versus 20# box). Also there was no way to know when data were missed. So in order to capture the no reads, a photoelectric sensor was used as a trigger. When the box moving on the conveyor belt breaks the photo sensor circ uit, it triggers the Linemaster to read data from both the RFID antenna and th e barcode reader. The photo sensors were programmed to trigger on rising signal as s hown in Figure 4-1. If the Linemaster doesn’t detect any RFID tag within some time range then it will time-out and send it as “NOREAD” to the software. This way Gato rPacker knows how many packer sticks may not have detected. No read data helps in a ssessing data collection performance. Figures 42 and 4-3 show the number of boxes scanned wi th and without triggers, respectively. The data for both graphs in Figures 4-2 and 43 were collected when the conditions were similar and the same types of items were packed. The average number of scans per day using a trigger was 3000±800. When the tri gger was not used, the average number of scans was 1900±700. This shows that about 1000 reads per day were missed when the trigger was not used. The only situation where the trigger was not effective was when production backups occurred on the conveyor belt, which causes boxes to touch each other. The reflective photo sensor requires a gap between boxes in orde r to detect a triggerable event. If the

PAGE 40

27 boxes on the conveyor belt are very close a nd touch each other then the system will detect it as a single box. Only one of the sticks put on the boxes are re ad. This problem is easily mitigated by progressively increasing sp eeds of conveyor belts. When boxes move from a slow moving conveyor to a fast movi ng conveyor they are automatically separated (Figures 4-4 and 4-5). Packer ID sticks re-distri bution and sorting was another issue. During the pilot, one person was exclusively reserved for collecting, sorting and re distributing RFID sticks to workers. To enable this person to sort quick ly, a metallic structure with many cylindrical slots was created. These cylindrical compar tments were numbered. RFID sticks were placed into their respective slots. The frequenc y that sticks needed to be redistributed was inversely related to the number of sticks per worker in th e system. To begin the pilot, each worker had 8 sticks. This was increased to 15 and then to 21, by the end of the pilot. Ultimately enough sticks to serve an 8 hour sh ift will likely be needed in order to minimize stick redistribution effort. The packer sticks had to withstand wet conditions, sanitization as well as mishandling from the workers. After two months 5% of the packer sticks were deemed to be damaged and unreadable. Since the tags we re “home-made” protot ypes they were not optimally designed for extended use in harsh conditions. Clear spray-on sealant was used to protect RFID labels from moisture. Alternat e methods can be used to make the sticks impervious to wet conditions. One approach ma y be to embed the tag inside the plastic itself. Figure 4-6 shows the condition of some tags after 2 months of use. Another issue that came up dur ing the pilot test was disc overing bad RFID tags. In order to discover which tags went bad, each one had to be manually inspected by using

PAGE 41

28 the RFID reader. This was a tedious process. In order to make this process efficient the Tag Management functionality was added to the software. The tag management feature generates a report specifying all tags that we re not seen within a time range specified by the user. For instance, if the time range sp ecified is from 8:00am -11:00am, then the report will list all issued tags that were not s een by the system within this interval. These data were useful because now only those tags that are shown in the report need to be checked. Tags not seen by the system for a given time were suspect of being damaged. Overall, data flow was smooth and consistent. Figure 4-7 shows the average unit productivity (number of scans/hour) of each p acker for the duration between June-15 and August 1. The graph indicates that packer number 20 was most productive at about 17 products per hour. Figure 4-8 shows the total nu mber of boxes packed on each day by all the packers. An economic analysis of implementing and using GatorPacker is presented in this section. Figure 4-7 represents the average productivity of each packer in terms of the number of scans. Average productivity of all the 19 packers was 10.83 scans/ hour and total productivity of all 19 packers wa s 205.84 scans/ hour. It is expected that implementation of GatorPacker will provide a means to increase worker productivity. As productivity increases there should be a need for fewer workers to achieve the same output. If it assumed that average individua l productivity can be increased to a level similar to the most productive workers, the number of workers that can be eliminated may be calculated. Since the cost of employing workers is fairly well known, estimates of costs versus benefits may be made. For exampl e, if it is assumed that average individual

PAGE 42

29 productivity is increased to a level equal to that of the av erage productivity of the top 5 workers of 14.31 scans/hour, an overall produc tivity of 205.85 scans/hour should be able to be accomplished with as few as 15 people. This provides a savings related to 4 workers. Costs to carry a single worker are about $14,000/year assuming workers earn about $6/hour, employment co sts are about $1/hour ($7/hr , 40 hours/week and 50 weeks per year). Therefore, a total approximate sa vings of $56,000 per year could be enjoyed. If this benefit were enjoyed for 5 years, this transl ates into an internal ra te of return of about 186% (Figure 4-9). However, it is likely that incentives will be required to achieve greater levels of productivity. If all remaining workers were offered a 15% increase ($0.90/hour), the benefit to the business would be about $29,000 per year for each similar production line, which translates into a payback period of about 1 year and a compelling internal rate of return of 93%. Figure 4-10 provi des a sensitivity analysis give n similar level of incentive, but with different numbers of workers eliminated.

PAGE 43

30 Figure 4-1. Rising Edge of Signal 0 500 1000 1500 2000 2500 3000 3500 4000 4500 12345678910111213141516 DayNumber of Scans Figure 4-2. Total Number of Boxes Scanned Per Day Using the Trigger

PAGE 44

31 0 500 1000 1500 2000 2500 3000 3500 4000 12345678910111213 DayNumber of Scans Figure 4-3. Total Number of Boxes Sca nned Per Day Without Using the Trigger Box 1 Conveyor Belt Box 2 Figure 4-4. Boxes Very Close to Each Other

PAGE 45

32 Figure 4-5. Changing the Speed of Conve yor Belts to Separate the Boxes Figure 4-6. Condition of Some Packer Tags After Two Months of Use

PAGE 46

33 0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 18.00201651815221263119984242111413Packer IDUnit Productivity(Scans/Hour) Unit Productivity(Scans/Hour) Figure 4-7. Overall Productivity Data of Each Packer between June 15 and August 1 0 500 1000 1500 2000 2500 3000 3500 4000 4500 12345678911121314151718192021DayNumber ofScans Figure 4-8. Total Number of Scans per Day.

PAGE 47

34 $30,000 1 0 2 345 $56,000 $56,000$56,000$56,000 $56,000 Figure 4-9. Cash Flow Diagram. 0%0% 34% 93% 148% 201% 0% 50% 100% 150% 200% 250% 123456 Workers RemovedIRR(%) Figure 4-10. IRR Sensitivity Analysis.

PAGE 48

35 CHAPTER 5 FUTURE WORK The pilot test has demonstrated promis e for the GatorPacker system. Additional work should be done to improve product id entification on the pr oduction line and to determine impacts of financial incentives, re al time reporting to workers, and other techniques on worker productivity. Specifica lly, the following items are planned for further development: Using Machine vision as an alternative to Barcodes and testing its effectiveness. Adding real time data ticker functionality to the admin console. This will provide the management with real time information about the packers Mounting computer monitors that would be linked to the software and viewed by workers and their supervisors. The monito rs would show metrics in real time, displaying how many boxes each packer packed. This could be combined with some type of reward system for the top five or 10 packers.

PAGE 49

36 LIST OF REFERENCES [1] R. A. Heinlein, The Moon Is a Harsh Mistress. New York: Tom Doherty Associates, 1966. [2] S. Garfinkel and B. Rosenberg, RFID: Applications, Security, and Privacy. Upper Saddle River, NJ: Addison Wesley, 2005. [3] Anonymous, “The History of RFID Technology,” RFID Journal , 20 Dec. 2005. http://www.rfidjournal.com/ar ticle/articleview/1338/1/129/ . Accessed Aug 2006 [4] R. Want, “The Magic of RFID ,” ACM Queue , vol. 2, no. 7, October 2004. http://www.acmqueue.com/modules.php?name=Content&pa=showpage&pid=216 Accessed Aug 2006 [5] EPCglobal Inc., “EPC Generation 1 Tag Data Standards Version 1.1 Rev.1.27,” 2005. http://www.epcglobalinc.org/standards/ . Accessed Aug 2006. [6] Microsoft Corporati on, “Technology Overview,” n.d. http://msdn.microsoft.com/netframewor k/technologyinfo/overview/default.aspx . Accessed Aug 2006.

PAGE 50

37 BIOGRAPHICAL SKETCH Karthik Thyagaraja received his Bachelor of Science degree in computer science and engineering from Visvesvaraya Nationa l Institute of Technology, India. He is currently pursuing his concurre nt masterÂ’s degrees in com puter science and engineering and agricultural and biologi cal engineering His academic interests include RFID, artificial intelligence and cognitive science. In his spare time he enjoys reading nonfiction science books and English classics.