Applying back propagation network to cold chain temperature monitoring

Recently, the enhancement of contactless, real-time features and high data transmission rates in supply chain management has been widely discussed. Notably, the cold chain is a part of the supply chain in which temperature monitoring plays a vital role in the system while automatic data collection is an essential aspect of cold chain management. As such, automatic data collection methods, such as radio frequency identification (RFID) techniques, are used to collect temperature data in order to track and trace all manner of products. In this paper, we evaluate an exponentially weighted moving average (EWMA) control chart and artificial neural network technologies in order to monitor collected temperature data in the context of cold chain management. The back-propagation neural network is used to predict temperature shifts and trends. The EWMA control chart is adopted to monitor temperature variations. Using this strategy, as anomalies occur, the control center of an enterprise can perform certain actions immediately to prevent further disaster. Finally, we construct a system using a back-propagation neural network and statistical process control chart. A simulated environment using LEGO(R) bricks is also implemented to demonstrate the feasibility of this study. This temperature control mechanism is found to be useful for a real-time temperature data collection environment, as with using active RFID tags with temperature sensors for cold chain management.

[1]  E. S. Page CONTINUOUS INSPECTION SCHEMES , 1954 .

[2]  James M. Lucas,et al.  Combined Shewhart-CUSUM Quality Control Schemes , 1982 .

[3]  A. Ovca,et al.  Maintaining a cold chain from purchase to the home and at home: consumer opinions. , 2009 .

[4]  Eugene L. Grant,et al.  Statistical Quality Control , 1946 .

[5]  Frédéric Thiesse,et al.  Sensor Applications in the Supply Chain: The Example of Quality-Based Issuing of Perishables , 2008, IOT.

[6]  S. W. Roberts Control chart tests based on geometric moving averages , 2000 .

[7]  Alberto Luceño,et al.  Choosing the EWMA Parameter in Engineering Process Control , 1995 .

[8]  James M. Lucas,et al.  Exponentially weighted moving average control schemes: Properties and enhancements , 1990 .

[9]  Ruey-Shiang Guh,et al.  Integrating artificial intelligence into on‐line statistical process control , 2003 .

[10]  Zheng Chen,et al.  A hybrid system for SPC concurrent pattern recognition , 2007, Adv. Eng. Informatics.

[11]  N F Hubele,et al.  X¯ CONTROL CHART PATTERN IDENTIFICATION THROUGH EFFICIENT OFF-LINE NEURAL NETWORK TRAINING , 1993 .

[12]  Shing I. Chang,et al.  A two-stage neural network approach for process variance change detection and classification , 1999 .

[13]  Lifeng Xi,et al.  A neural network ensemble-based model for on-line monitoring and diagnosis of out-of-control signals in multivariate manufacturing processes , 2009, Expert Syst. Appl..

[14]  J. P. Emond RFID Technology and Applications: The cold chain , 2008 .

[15]  James M. Lucas,et al.  A Modified "V" Mask Control Schemet , 1973 .

[16]  I A Basheer,et al.  Artificial neural networks: fundamentals, computing, design, and application. , 2000, Journal of microbiological methods.

[17]  J. R. Tony Arnold,et al.  Introduction to Materials Management , 1991 .

[18]  P. Werbos,et al.  Beyond Regression : "New Tools for Prediction and Analysis in the Behavioral Sciences , 1974 .

[19]  Ronald B. Crosier,et al.  A new two-sided cumulative sum quality control scheme , 1986 .

[20]  Chuen-Sheng Cheng A multi-layer neural network model for detecting changes in the process mean , 1995 .

[21]  Charles W. Champ,et al.  Exact results for shewhart control charts with supplementary runs rules , 1987 .