Boyang Song
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Boyang Song has a total of 9 co-authors. Here are the top co-authors by number of shared publications:
- Ashutosh Tiwari
Cranfield University
2 shared publications Vinayak Ashok Prabhu
Manufacturing Informatics Centre, Cranfield University, Bedfordshire, UK
2 shared publications- Ashutosh Tiwari
School of Mechanical, Aerospace and Civil Engineering, The University of Sheffield, Sheffield, U.K.
1 shared publication - Divya Tiwari
School of Mechanical, Aerospace and Civil Engineering, The University of Sheffield, Sheffield, U.K.
1 shared publication - John Thrower
Centre for Advanced Systems, Cranfield University, Cranfield, UK
1 shared publication - Michael Farnsworth
Department of Automatic Control and Systems Engineering, The University of Sheffield, Sheffield, U.K.
1 shared publication
Latest Publications
- Open Access
- 1 July 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Sensors Journal
- Vol. 22 (15) , 15235-15247
Deep Transfer Learning With Self-Attention for Industry Sensor Fusion Tasks
Monitoring of complex industrial processes can be achieved by obtaining process data by utilising various sensing modalities. The recent emergence of deep learning provides a new routine for processing multi-sensor information. However, the learning ability of shallow neural networks is insufficient, and the data amount required by deep networks is often too large for industrial scenarios. This paper provides a novel deep transfer learning method as a possible solution that offers an advantage of better learning ability of the deep network without the requirement for a large amount of training data. This paper presents how Transformer with self-attention trained from natural language can be transferred to the sensor fusion task. Our proposed method is tested on 3 datasets: condition monitoring of a hydraulic system, bearing, and gearbox dataset. The results show that the Transformer trained from natural language can effectively reduce the required data amount for using deep learning in industrial sensor fusion with high prediction accuracy. The difficult and uncertain artificial feature engineering which requires a large workload can also be eliminated, as the deep networks are able to extract features automatically. In addition, the self-attention mechanism of Transformer aids in the identification of critical sensors, hence the interpretability of deep learning in industrial sensor fusion can be improved.
- 1 July 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Sensors Journal
- Vol. 22 (15) , 15235-15247
Open AccessDeep Transfer Learning With Self-Attention for Industry Sensor Fusion Tasks
Monitoring of complex industrial processes can be achieved by obtaining process data by utilising various sensing modalities. The recent emergence of deep learning provides a new routine for processing multi-sensor information. However, the learning ability of shallow neural networks is insufficient, and the data amount required by deep networks is often too large for industrial scenarios. This paper provides a novel deep transfer learning method as a possible solution that offers an advantage of better learning ability of the deep network without the requirement for a large amount of training data. This paper presents how Transformer with self-attention trained from natural language can be transferred to the sensor fusion task. Our proposed method is tested on 3 datasets: condition monitoring of a hydraulic system, bearing, and gearbox dataset. The results show that the Transformer trained from natural language can effectively reduce the required data amount for using deep learning in industrial sensor fusion with high prediction accuracy. The difficult and uncertain artificial feature engineering which requires a large workload can also be eliminated, as the deep networks are able to extract features automatically. In addition, the self-attention mechanism of Transformer aids in the identification of critical sensors, hence the interpretability of deep learning in industrial sensor fusion can be improved.
- 16 May 2016
- Published by SAGE Publications in Proceedings of the Institution of Mechanical Engineers, Part B: Journal of Engineering Manufacture
- Vol. 232 (1) , 51-66
Decision support system enabled by depth imaging sensor data for intelligent automation of moving assemblies
Automation, driven by informatics, enables manufacturing companies to increase productivity and meet market demands for cost-effective and high-quality products. However, many manufacturing operations across industry verticals continue to be manual even today. One such example is the manual assembly of the final trim and wheels in an automotive production line where there is heavy reliance on human decision-making pertaining to when, where and how to install components on and inside a constantly moving vehicle body. The main aim of this work is to develop a rule-based decision support system that will enable an automation solution to make human-like decisions in moving assembly operations. The wheel loading operation is chosen as a case study and a decision support framework and tool is developed and successfully tested using multiple assembly scenarios generated from experimental data provided by gaming interface sensors installed on the laboratory-based shopfloor. The resulting decision support system has the potential to enable the automation of moving assembly operations in various sectors of the manufacturing industry.
- 16 May 2016
- Published by SAGE Publications in Proceedings of the Institution of Mechanical Engineers, Part B: Journal of Engineering Manufacture
- Vol. 232 (1) , 51-66
Decision support system enabled by depth imaging sensor data for intelligent automation of moving assemblies
Automation, driven by informatics, enables manufacturing companies to increase productivity and meet market demands for cost-effective and high-quality products. However, many manufacturing operations across industry verticals continue to be manual even today. One such example is the manual assembly of the final trim and wheels in an automotive production line where there is heavy reliance on human decision-making pertaining to when, where and how to install components on and inside a constantly moving vehicle body. The main aim of this work is to develop a rule-based decision support system that will enable an automation solution to make human-like decisions in moving assembly operations. The wheel loading operation is chosen as a case study and a decision support framework and tool is developed and successfully tested using multiple assembly scenarios generated from experimental data provided by gaming interface sensors installed on the laboratory-based shopfloor. The resulting decision support system has the potential to enable the automation of moving assembly operations in various sectors of the manufacturing industry.
- Open Access
- 9 October 2015
- Published by Springer Nature in The International Journal of Advanced Manufacturing Technology
- Vol. 85 (1-4) , 163-184
Digitisation of a moving assembly operation using multiple depth imaging sensors
- 9 October 2015
- Published by Springer Nature in The International Journal of Advanced Manufacturing Technology
- Vol. 85 (1-4) , 163-184
Open AccessDigitisation of a moving assembly operation using multiple depth imaging sensors