Zeng Long

2007-2012: Ph.D., CAD/CAM, Department of Mechanical and Aerospace Engineering, Hong Kong University of Science and Technology

2005-2007: MPhil, Mechanical design and theory, Zhejiang University

2001-2005: Bachelor's, Measurement and control technology and instruments, Dalian University of Technology

2014-present: Assistant Professor, Division of Advanced Manufacturing, Graduate School at Shenzhen, Tsinghua University

2012-2014: Research associate, Department of Mechanical and Aerospace Engineering, Hong Kong University of Science and Technology

Product Design and Development; Advanced Mechanical Engineering Design; Modern CAD Methodology and Technology

 Intelligent design; Computer-aided design

Mobile device oriented Intelligent CAD techniques

As the touch-based input method is adopted widely by mobile devices in recent years, e.g. phones and pads, designers have put forward intelligent demands for the CAD system based on mobile devices, to overcome three major challenges. First, the iDesignCAD system should understand users’ design intents correctly and perform in a reasonable way, because we cannot tell system what we want to do via traditional menus. Second, the input data from touch-based input is inaccurate, in contrast, the data of engineering models is precise. Third, the algorithms and data structure should be efficient since the computational power and storage space of mobile devices are limited.

Vision-guided robot control and applications

Intelligent flexible assembly is designed for small amount customization. As the figure shown, parts are usually delivered to the assembly center after they are manufactured. The intelligent flexible assembly system has three major benefits: 1) an assembly line can assemble several types of products; 2) it can save feeding and unloading devices since no strict requirements on parts’ postures; 3) it is easy and low-cost in switching among products.

Vision-guided robot is a critical and fundamental technique. Vision for robot is just like eyes for human, while algorithms like brain. It determines a robot’s intelligent level. It can not only be applied to flexible assembly, but also to service robots, e.g. home-care area.   

On-machine inspection system for 5-axis CNC 

OMI is an inspection system by installing a probe head onto a CNC.  It can monitor and inspect a manufacturing process. It avoids errors from second-clamping, for precision and large-scale structural parts. The figures shows large thin-wall structural parts of aircraft. A second-clamping will cause unacceptable deformation error.

An intelligent OMI system should inspect each point accurately. In addition, it can generate inspection points as few as possible, plan an optimal inspection path automatically without interference, and simulate inspection process. Furthermore, it can compute various accuracy, e.g. flatness, and perform trend analysis from an amount of inspection data, which can predict CNC status, such as cutting-tool wear, can adjust manufacturing processes adaptively.

• Sketch-based assembly modeling with intelligent features, no. 61502263, NSFC, 249,000RMB, 2016-2018.

• Research on flexible assembly techniques, IPE Group Limited, 350,000RMB, 2017-2018.

• Intelligent on-machine inspection system development,  Zhejiang RIFA Ltd., 2,040,000RMB，2018-2019.

1. Li B, Feng P, Zeng L, et al. Path planning method for on-machine inspection of aerospace structures based on adjacent feature graph[J]. Robotics and Computer-Integrated Manufacturing, 2018, 54:17-34.

2. Mi S L, Wu X Y, Zeng L. Optimal build orientation based on material changes for FGM parts[J]. International Journal of Advanced Manufacturing Technology, 2017, 94(3):1-14.

3. Xu Y, Fan T, Xu M, Zeng L. SpiderCNN: Deep Learning on Point Sets with Parameterized Convolutional Filters, ECCV, 2018.

4. L. Zeng, Y. J. Liu, J. Wang, D. L. Zhang, and M. M. F. Yuen. Sketch2Jewelry: Semantic Feature Modeling for Sketch-based Jewelry Design, Computers & Graphics (JCR二区), Vol.38, pp.69-77, 2013.

5. L. Zeng, Y. J. Liu, M. Chen, and M. M. F. Yuen. Least Squares Quasi-Developable Mesh Approximation, Computer Aided Geometric Design (JCR二区), Vol.29, No. 7, pp.565-578, 2012.

6. L. Zeng, Y. J. Liu, S. H. Lee, and M. M. F. Yuen. Q-Complex: Efficient Non-Manifold Boundary Representation with Inclusion Topology, Computer-Aided Design (JCR一区), Vol. 44, No. 11, pp.1115-1126, 2012.

7. L. Zeng, L. M. L. Lai, D. Qi, Y. H. Lai, M. M. F. Yuen. Efficient Slicing Procedure based on Adaptive Layer Depth Normal Image, Computer-Aided Design (JCR一区), Vol.43, No. 12, pp.1577-1586, 2011.

8. L. Zeng, J. Wang, Y. J. Liu, and M. M. F. Yuen. Topological Data Transfer among Explicit Non-Manifold B-reps for Polygonal Models (Poster). GMP, June18-24, 2012, Huangshan, China.

9. L. Zeng, M. Chen and M. M. F. Yuen. Developable Mesh Surface Approximation by Normal Guided Deformation (short paper). In: proceedings of Computer Graphics International Conference, June 8-11, 2010, Singapore.

10. D. Qi, L. Zeng, M. M. F. Yuen. Robust Slicing Procedure based on Surfel-Grid, Computer-Aided Design and Applications, pp. 965-981,Vol.10, No. 6, 2013.

11. J. Y. Jia, Q. Zhang, L. Zeng, S. Liang. Voxel-Encoded Descriptor for 3D Model Retrieval by Exploring Model’s Spatial Information, Journal of Mechanical Science & Technology (JCR三区), Vol.28(7),  2014.