Adoption of RFID technology in warehouse

Adoption of RFID technology in warehouse management MODRÁK, Vladimír - KNUTH, Peter - ŠEBEJ, Peter

Abstract. RFID technology itself can be beneficial in many areas, including warehouse systems. This article deals with optimal tag positioning on the goods, secondly on the practical knowledge from testing tags readability in a real situation and finally it is focused on setup and testing of portal solution for inbound and outbound goods detection. Measurements and findings are not only described, but are also demonstrated by graphs and comprehensively analyzed. Keywords: RFID, UHF, testing, warehouse, management system.

1 Introduction No one doubts that RFID technology rapidly and beneficially penetrates into our day life and industrial applications. In spite of that, bar code technology is still holding its dominant position in small and medium enterprises (SMEs). On the other hand RFID technology gradually finds its position also in SMEs. One of the potential areas where this technology can be effectively exploited is warehouse operation. In this field, RFID can be used as a tool for achieving real-time state of the warehouse and use this information for further use in planning, controlling, etc. The strong motivation for this effort is a fact that automation of data collection by this mean will require less manpower and will increase reliability of warehouse operations. Furthermore, in compare with bar code technology, the process of data entry will speed up several times. These benefits are not possible without direct connection between RFID technology and warehouse information system through interface. This paper deals mainly with three areas. Firstly, identification of the optimal tag position in the environment of warehouse with a plastic raw material, secondly on the practical knowledge from testing tags readability in a real situation and finally setup and testing of portal solution for inbound and outbound goods detection.

2 Related work According to Gu [3], warehouse information system and management has a big role in linking the supply chain partners and making smooth flow of goods inside the

chain. Therefore handling resources and units in the warehouse is essential for smooth production operation, inventory and transshipment costs reduction, and productivity improvement in the factory [14]. Nowadays, common information systems such as MES/ERP systems or/and warehouse management systems can not have an access to detailed information and they have no idea of what is really happening to material flow on the shop floor. [8]. Visualization of actual working status and real-time information gathering is lacking in these systems too [4]. Furthermore such systems require human interaction in form of manual based or bar-code based data entry and therefore human factor seems to be the weakest point [10]. Intelligent systems, as RFID technology can be used to eliminate human errors and provide real-time automated process of data gathering. RFID solutions support a larger set of unique IDs than bar codes and can incorporate additional data such as manufacturer, product type, and even measure environmental factors such as temperature [15]. Enhanced productivity in the warehouse by use of RFID technology and logistics resource management systems is mainly achieved by reduction of time for storing and retrieving data to the warehouse information system [9]. Furthermore warehouses that use RFID based management system can significantly reduce amount of pallets and truck wailings [16]. Even big companies as Wal-Mar are using RFID technology in their supply chain and warehouse management, which gives them and their suppliers’ significant benefits [12]. Liu [6] used this technology to solve manual operational mistakes in the production process and developed RFID material flow control system by use of RosettaNet network and ERP system. Results showed benefits as increased productivity, less work load and incensement of information speed and accuracy. Also Vemuri [13] stresses importance and advantages of RFID technology as a solution for inventory inaccuracies, which does not need line-of-sight, and is able to track and manage inventory items as a tool of the warehouse information system. Logistic resource management system for material handling solutions in the warehouse was proposed by Poon et.al. [9]. He describes possible way of real-time data collection and sharing in the warehouse together with an effective triangular localization scheme. Furthermore reading performances of two different frequencies and types (active and passive) on the various materials have been tested. This example was tested and demonstrated under real conditions in Group Sense Limited. Design and operation warehouse problems are more described by Gu et.al. [3]. Even though initial investments into RFID technology is higher than in case of bar-code systems, companies are willing to implement it in order to improve the data collection [7]. Gonzales [2] made this year significant research on warehousing and data mining of massive RFID sets and extended his research of proposed RFID-cuboid concept. Extension is mainly focused on creation of more realistic account of item movement model. Even if a lot of work has been done in this field, further testing and investigation is still needed.

3 Determination and adoption of RFID principles The first step was determination of suitable technology for recognizing goods in the plastic raw material warehouse. In generally, three basic principles of RFID tags could be used [1],[15]: - active RFID tags (contain battery, therefore they can be read from longer distances), - passive RFID tags (tags are supplied by energy transmitted by RFID reader), - semi-passive RFID tags (combination of the above tags) All three types of above mentioned tags operate on the various frequencies, mostly known as [5]: - LF → less than 135 kHz → mostly passive, biggest antenna of all RFID tags, no or poor anti-collision algorithm, - HF → 13,56 MHz → mostly passive, cheaper than LF RFID tags, used in Smart Cards, - UHF → 860 - 930 MHz → in large volumes cheaper than LF and HF RFID tags, longer read distance than LF and HF, good readability of multiple tags, - Microwave → 2,45 GHz → similar properties as UHF RFID tags, but with higher read speed, typically used in electronic toll systems or location tracking. In this research passive UHF RFID tag technology was selected as the most suitable for use in warehouse with plastic raw material. For this purpose Alien ALR8800 that operates on 866 MHz was identified as the applicable solution, mainly because of new features enabling identification of the tag velocity and direction supported directly by the reader’s firmware. In order to determine radiation diagram and optimal tag orientation of the Alien reader, measurements have been carried out in free room with one standard antenna and Alien ALL-9540-02 passive tag (called also world tag because it can operate from 860 MHz to 960 MHz). Reader was connected by serial output and consequent USB reduction to the USB port of the computer. Alien RFID Gateway software has been used to identify the interrogation zone of the tag that was attached to the wooden stick. Passive tag and antenna were placed in the same altitude and reader’s output signal strength was set to maximum. Surrounding electromagnetic field was measured to ensure that they will not influence the measurement. Overall, three different positions of the tag towards the RFID antenna have been tested: 1. horizontally towards floor and parallel to RFID antenna 2. vertically towards floor and parallel to the RFID antenna 3. vertically towards floor and perpendicular to the RFID antenna Other positions of the tag were unreasonable, because the tag can not receive enough energy to send back information requested by the reader if placed only by its edge towards the RFID reader antenna. Therefore horizontally placed tag to the RFID antenna is unreadable and does not require further testing of this position. In each position, the tag was moved through the empty room from the close front of the antenna. The room was empty in order to minimize influence of the surrounding materials to the tag/reader communication, because in these communication frequencies the waves are not only absorbed, but also reflected. In the event that the tag was not recognized anymore by the reader, the border position of the tag read range was recorded. Summary measurement results are shown on the Fig.

1. The graphical grid on the Fig. 1 represents 1 meter lengths and each line shows different tag position as mentioned above. Rectangle displays Alien’s RFID reader antenna. Specification behind the antenna was not measured, because for given purpose it wasn‘t considered as relevant.

Tag position horizontally towards floor and parallel to the RFID antenna Tag position vertically towards floor and parallel to the RFID antenna Tag position vertically towards floor and perpendicular to the RFID antenna

Fig. 1. Interrogation zone of ALR-8800 antenna.

Based on the measurement characteristics shown on the Fig. 1, it can be determined that maximal read range in these particular conditions is 6,5m. Measurements have also shown how to place the tag if we need to achieve maximal or minimal read range. According to Fig. 1 it is obvious that highest read range can by achieved by placing tag vertically towards floor and parallel to the RFID antenna and minimum read range can be achieved by positioning tag horizontally towards floor and parallel to the RFID antenna. This is essential, because not all applications of RFID technology require just the maximal read range. Some applications like RFID terminals for items identification demand to have accurate read ranges, but not higher than requested.

3.1 RFID technology adoption Verification of optimal tag position was carried out in under real conditions in company producing plastic components. Based on the company’s demands, the RFID implementation was intended to be involved in the modernization of their warehouse management and, especially with focus to real-time data collection to the warehouse information system. This meant to begin implementation process by mapping all information flows and specific material handling conditions in the warehouse. Initial state of the warehouse organizing and management can be described as following. After receiving of raw material, which is mainly in the form of granulates, items are stored and manually registered into warehouse information system. Each type of granulate has its unique code, which is subsequently used for raw material (item) identification. Releasing of items into buffer stock is organized by FIFO. Overall, warehouse contains around 60 to 70 kinds of material items. Buffer stock is planned to be sufficient for 24 hours. The requirement to know actual warehouse state is related to optimization of warehouse stock in order to decrease financial sources lock-up. As output items from production are labeled by barcode, company managers preferred solution based on this technology. Due to the fact of mentioned disadvantages of barcodes, we decided to verify for automation purpose UHF RFID technology. To test suitability of this technology it was necessary to: - Determine where RFID antenna(s) will be situated. Since fork lift is in fact continuously moving within the warehouse, it was decided to try place antenna directly to the fork lift. Moreover it was necessary to locate antenna into checkin and check-out gates. Due to the needs to cover the size of check-in and checkout gates it would be necessary to install four antennas (see Fig. 2). - Simulate situation in which tags can be located on any place in the warehouse. Therefore tags were placed on critical locations and places in the warehouse. - Carry out testing by use of our laboratory RFID technology configuration (see Fig. 3). During the testing fork lift was performing ordinal manipulation activities and simultaneously served as carrier of RFID reader, antenna and wireless access point. - Design and test input-output portal for identification of material flow into and out of the warehouse.

Fig. 2. Interrogation zone of antennas.

In order to determine a suitable position of antennas on the forklift, a hand pallet truck with the RFID configuration shown on the Fig. 3 was used. Possible surrounding sources of other electromagnetic fields have been checked by electromagnetic field meter and no sources that could interference the reader have been found. As it can be seen from the Fig. 3, reader is in this case directly connected to the computer through telnet-command Alien RFID Gateway software. This enabled us to determine by repetitive measurements in various altitudes that the best results were achieved when antenna was 0,7m above the floor.

Fig. 3. RFID reader Alien ALR-8800 with antenna and running software.

During measurements we had on-line information about the tag unique identification number, its direction, velocity and distance from the reader as it is shown on the Fig. 4.

Fig. 4. Print screen of identified RFID tags with velocity and distance value.

After mounting the whole configuration on the forklift, two antennas were attached on the sides of the forklift and two antennas on the front. Front antennas were intended for reading of items loaded on the fork lift. In addition, wireless access point was attached to the reader in order to enable free movement of the forklift. The response time of the reader was checked to ensure that wireless connection is working properly and that packets are reaching its destination. Wireless connection was secured by WPA2 and AES key. Overall results of measurements are shown on the Fig. 5. The unique number of each measured tag begins with E200 3411 B802 0114 1225 55xx, where xx represents number in the legend on the Fig. 5.

8

Visualization of tag reads during fork lift cruise through the warehouse i17

7

i18

6

Tag number

i22

5

i23

4

i24

3

i27

2

i60

1 0

Time of the event [hh:mm:ss,ms]

Fig. 5. Graph of measurement results showing tag reads in the time.

Testing of RFID system was based on the following principle. By moving of antenna, which was situated on the fork lift were one by one recorded events about tag read and tag signal loss. During experiment we recorded 127 events. Simulation of real conditions took 238 seconds. This time was in our opinion sufficient, because during this period fork lift one time repeated common trajectory that simulated real work conditions in the warehouse. Frequency of the tag read was set to 1 second. Tag was positioned vertically towards floor and parallel to the RFID antenna in order to achieve maximal read range. The final measurement results shown on the Fig. 5 are based on the following rules (R): R1: In one event can be just one tag read. R2: Two or more neighbor events can occur in the same time. Results of measurements during fork lift cruise through the warehouse showed that all tags were read repeatedly but in different appearance. To conclude obtained measurement results it is possible to formulate the following findings: 1. Repetitive read occurrence of each tag during given time period at least adequately if not supernumerary suits requirements for operative evidence of stock items in real time regime. 2. Different appearances of tag reads during the same time period were caused by non-equal distances and surrounding environment. Based on the above findings, it can be stated that all tags have been read correctly and tested technology was proved to be suitable for the given conditions and surrounding materials. Subsequently, after verification of the RFID components configuration on the forklift, we decided to test the portal-based RFID configuration. Therefore we attached two antennas connected to the RFID reader on the one side of the warehouse entrance. RFID reader was set to read the tags every second and was connected wirelessly to the computer with RFID communication software. In order to verify interrogation zones, items on the pallet have been labeled and moved through all possible critical positions. This included both edges of the entrance, lower and upper side of the warehouse entrance. For carrying of the items, pallet truck instead of forklift was used (see Fig. 6). Information such as direction and velocity of the tags moving through the warehouse entrance has been recorded into the log file. Pallet truck was moved for six minutes into and out of the warehouse (as outlined on the Fig. 6) to simulate material

flow. During this period the reader recorded total 686 events. Another 2803 reads have been made in the lower edges of the entrance and 837 reads in the upper side of the warehouse entrance.

1

Legend:

4 5

3

2 6 7

4

1- RFID antenna connected to the reader 2- interrogation zone 3- pallet truck 4- pallet 5- tag E200 3411 B802 0114 1225 5517 6- tag E200 3411 B802 0114 1225 5518 7- direction of movement

Fig. 6. Layout of testing experiment in the warehouse.

As it can be seen from the fragment of recorded data shown on the graphs (Fig. 7 and Fig. 8), tags, even if positioned not far from each other (see Fig. 6), have different measured velocity, which was depended on the distance of the tags from the reader’s antenna. Anti-collision algorithm also played its role in these results, because tags are not read at once, but one by one. But after all, the direction of tag movement was recognized correctly ((+) represents inbound items and (–) outbound items). Accordingly, information about the tag movement direction can be beneficially used for real-time automatic data entry, which reflects real material flow into and out of the warehouse.

Velocity [m/s]

Tag E200 3411 B802 0114 1225 5517 0,8 0,6 0,4 0,2 0 -0,2 -0,4 -0,6

Time of the event [hh:mm:ss,ms]

Fig. 7. Graph of E200 3411 B802 0114 1225 5517 tag readings and velocity.

Velocity [m/s]

Tag E200 3411 B802 0114 1225 5518 0,5 0,4 0,3 0,2 0,1 0 -0,1 -0,2 -0,3 -0,4 -0,5

Time of the event [hh:mm:ss,ms]

Fig. 8. Graph of E200 3411 B802 0114 1225 5517 tag readings and velocity.

If we take closer look to the graph on the Fig. 8, we can see that the density of readings is much higher in compare with the tag displayed on the Fig. 7. The probable reason of this effect is caused by characteristic of the antenna interrogation zone and overall anti-collision reading procedure of the reader.

4 Results and conclusion In generally, an adoption of RFID technology in warehouse management is due to many specifics rather long-term problem. This paper describes testing and preparation stages of the pilot project, in which only passive RFID technology components were applied. According to above mentioned practical experiences it is possible to say that selected RFID principle and components are suitable for effective use in the given conditions. Moreover, it was successfully verified that tag movement parameters can be effectively used for recording of altered amount of goods in the warehouse. Materials and items stored in the warehouse didn’t cause considerable signal loss and tags were responding timely even at higher distances. Evenly, tags were responding timely if they were placed in the edge of the warehouse and surrounded by other pallets with raw material. In a further investigation, cost effective analysis of proposed solution could be carried out to convince managers and stakeholders to make investment to this advanced technology.

References 1. Finkenzeller, K.: RFID-Handbook, 2nd Edition: Fundamental and Applications in Contactless Smart Cards and Identification, Wiley & Sons Ltd., Swadlincote (2003) 2. Gonzalez, H., Han, J., X. Li, X., Klabjan, D.: Warehousing and Analysis of Massive RFID Data sets, Proceedings ICDE 2006, (2006)

3. Gu, J., Goetschalckx, M., & McGinnis, L. F.: Research on warehouse operation: A comprehensive review. European Journal of Operational Research, 177, pp. 1–21 (2007) 4. Huang, G. Q., Zhang, Y. F., & Jiang, P. Y.: RFID-based wireless manufacturing for walking-worker assembly islands with fixed-position layouts. Robotics and ComputerIntegrated Manufacturing, 23(4), pp. 469–477 (2007) 5. Lewis, S.: A basic introduction to RFID technology and its use in the supply chain, White Paper (2005) 6. Liu, C. M.: An electronic material flow control system for improving production efficiency in integrated-circuit packaging industry. In Proceedings of IEEE international conference on RFID Eurasia, pp. 1–8 (2007) 7. Morrison, J.: Help wanted. RFID Journal, 13–20 (2005) 8. Modrák, V., Mandulák, J.: Mapping Development of MES Functionalities. ICINCO 2009, Proceedings 3 SPSMC, pp. 244-247 (2009) 9. Poon, T. C., Choy, K. L., Chow, H. K. H., Lau, H. C. W., Chan, F. T. S., & Ho, K. C.: A RFID case-based logistics resource management system for managing order- picking operations in warehouses. Expert Systems with Applications, 36, pp. 8277–8301 (2009) 10. Sexton, J. B., Thomas, E. J., & Helmreich, R. L.: Error, stress, and teamwork in medicine and aviation: Cross sectional surveys. BMJ2000, 320(7237), pp. 745–749 (2000) 11. Tan, T.-H., & Chang, C.-S.: Development and evaluation of an RFID-based e-restaurant system for customer-centric service. Expert Systems with Applications (2010) 12. Tajima, M.: Strategic value of RFID in supply chain management. Journal of Purchasing and Supply Management, 13(4), pp. 261–273 (2007) 13. Vemuri, K.V.: The benefits of accurate and timely data lean in production environments, In: Proceedings of the Eighth ICEIS 2006, Pathos, Cyprus, May 23-27, pp. 221-225 (2006) 14. Vogt, J.J., Pienaar, W.J., Wit, P.W.C. De: Business logistics management: Theory and practice, Oxford University Press, Oxford (2005) 15. Want, R.: An introduction to RFID technology. Pervasive Computing, IEEE, Vol. 5 (1), pp. 25 – 33. (2006) 16. Modrák, V., Knuth, P., Novak-Marcincin, J.: Advantages and risks of RFID in business applications. International business management 4 (1), pp. 28-34 (2010)