Deep neural networks based order completion time prediction by using real-time job shop RFID data

In the traditional order completion time (OCT) prediction methods, some mutable and ideal production data (e.g., the arrival time of work in process (WIP), the planned processing time of all operations, and the expected waiting time per operation) are often used. Thus, the prediction time always deviates from the actual completion time dramatically even though the dynamicity of the production capacity and the real-time load conditions of job shop are considered in the OCT prediction method. On account of this, a new prediction method of OCT using the composition of order and real-time job shop RFID data is proposed in this article. It applies accurate RFID data to depict the real-time load conditions of job shop, and attempts to mine the mapping relationship between RFID data and OCT from historical data. Firstly, RFID devices capture the types and waiting list information of all WIPs which are in the in-stocks and out-stocks of machining workstations, and the real-time processing progress of all WIPs which are under machining at machining workstations. Secondly, a description model of real-time job shop load conditions is put forward by using the RFID data. Next, the mapping model based on the composition of order and real-time RFID data is established. Finally, deep belief network, which is one of the major technologies of deep neural networks, is applied to mine the mapping relationship. To illustrate the advantages of the proposed method, a numerical experiment compared with back-propagation (BP) network based prediction method, multi-hidden-layers BP network based prediction method and the principal components analysis and BP network based prediction method is conducted at last.

[1]  Christian Szegedy,et al.  DeepPose: Human Pose Estimation via Deep Neural Networks , 2013, 2014 IEEE Conference on Computer Vision and Pattern Recognition.

[2]  Le Gruenwald,et al.  Real-time due-date promising by build-to-order environments , 2004 .

[3]  feng liang.  A Combined Approach of Cycle Time Estimation in Mass Customization Enterprise , 2016 .

[4]  Shao-Chung Hsu,et al.  Due date assignment using artificial neural networks under different shop floor control strategies , 2004 .

[5]  S. T. Enns,et al.  A dynamic forecasting model for job shop flowtime prediction and tardiness control , 1995 .

[6]  Yoshua Bengio,et al.  Greedy Layer-Wise Training of Deep Networks , 2006, NIPS.

[7]  Geoffrey E. Hinton.  Training Products of Experts by Minimizing Contrastive Divergence , 2002, Neural Computation.

[8]  V. Vinod,et al.  Simulation modeling and analysis of due-date assignment methods and scheduling decision rules in a dynamic job shop production system , 2011 .

[9]  Geoffrey E. Hinton,et al.  Acoustic Modeling Using Deep Belief Networks , 2012, IEEE Transactions on Audio, Speech, and Language Processing.

[10]  Xin Zheng,et al.  Application of deep neural network in estimation of the weld bead parameters , 2015, 2015 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS).

[11]  Lars Mönch,et al.  Incorporating engineering process improvement activities into production planning formulations using a large-scale wafer fab model , 2016 .

[12]  Yun Zhang,et al.  Analysis of Feature Extracting Ability for Cutting State Monitoring Using Deep Belief Networks , 2015 .

[13]  Stephen R. Lawrence,et al.  Estimating flowtimes and setting due-dates in complex production systems , 1995 .

[14]  Donatella Corti,et al.  Due date (DD) quotation and capacity planning inmake-to-order companies: results from an empirical analysis , 2008 .

[15]  Honglak Lee,et al.  Convolutional deep belief networks for scalable unsupervised learning of hierarchical representations , 2009, ICML '09.

[16]  Wallace J. Hopp,et al.  A Simple, Robust Leadtime-Quoting Policy , 2001, Manuf. Serv. Oper. Manag..

[17]  Z. K. Weng,et al.  Manufacturing lead times, system utilization rates and lead-time-related demand , 1996 .

[18]  Geoffrey E. Hinton,et al.  Reducing the Dimensionality of Data with Neural Networks , 2006, Science.

[19]  Pingyu Jiang,et al.  Manifold learning based rescheduling decision mechanism for recessive disturbances in RFID-driven job shops , 2016, Journal of Intelligent Manufacturing.

[20]  Donatella Corti,et al.  A capacity-driven approach to establish reliable due dates in a MTO environment , 2006 .

[21]  Haibo Shi,et al.  Application of support vector regression in prediction of due date under uncertain assemble-to-order environment: Application of support vector regression in prediction of due date under uncertain assemble-to-order environment , 2013 .

[22]  Tao Zhang,et al.  Estimation of Lead Time in the RFID-Enabled Real-Time Shopfloor Production with a Data Mining Model , 2013 .

[23]  Zhang Xiangwei,et al.  Order Completion Date Estimation and Due Date Decision under Make-to-Order Mode , 2012 .

[24]  Noel Lopes,et al.  Deep Belief Networks (DBNs) , 2015 .

[25]  Ming Li,et al.  Due date assignment and dynamic scheduling of one-of-a-kind assembly production with uncertain processing time , 2015, Int. J. Comput. Integr. Manuf..

[26]  Yee Whye Teh,et al.  A Fast Learning Algorithm for Deep Belief Nets , 2006, Neural Computation.

[27]  Ihsan Sabuncuoglu,et al.  Operation-based flowtime estimation in a dynamic job shop , 2002 .

[28]  Vitaly A. Strusevich,et al.  Earliness penalties on a single machine subject to precedence constraints: SLK due date assignment , 1997, Comput. Oper. Res..

[29]  Ali Azadeh,et al.  A Neuro-Fuzzy-Regression Algorithm for Improved Prediction of Manufacturing Lead Time with Machine Breakdowns , 2011, Concurr. Eng. Res. Appl..

[30]  Pingfeng Wang,et al.  Failure diagnosis using deep belief learning based health state classification , 2013, Reliab. Eng. Syst. Saf..

[31]  Suh-Jenq Yang,et al.  Multiple common due dates assignment and scheduling problems with resource allocation and general position-dependent deterioration effect , 2013 .

[32]  Toly Chen.  A hybrid fuzzy-neural approach to job completion time prediction in a semiconductor fabrication factory , 2008, Neurocomputing.

[33]  El-Houssaine Aghezzaf,et al.  Integrating order acceptance decisions with flexible due dates in a production planning model with load-dependent lead times , 2015 .

[34]  Chengbin Chu,et al.  A survey of the state-of-the-art of common due date assignment and scheduling research , 2002, Eur. J. Oper. Res..

[35]  Bhuvana Ramabhadran,et al.  Deep belief nets for natural language call-routing , 2011, 2011 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[36]  Toly Chen,et al.  Incorporating fuzzy c-means and a back-propagation network ensemble to job completion time prediction in a semiconductor fabrication factory , 2007, Fuzzy Sets Syst..