Wenxing Liu

Dr. Wenxing Liu

orcid

Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
19 July 2022 in IEEE Transactions on Neural Networks and Learning Systems
Verified
Journal Article
Closed access
WLWenxing Liu
Wenxing Liu
HNHanlin Niu
Hanlin Niu
IJInmo Jang
Inmo Jang
GHGuido Herrmann
Guido Herrmann
JCJoaquin Carrasco
Joaquin Carrasco
In this article, we propose an algorithm that combines actor–critic-based off-policy method with consensus-based distributed training to deal with multiagent deep reinforcement learning problems. Specifically, convergence analysis of a consensus algorithm for a type of nonlinear system with a Lyapunov method is developed, and we use this result to analyze the convergence properties of the actor training parameters and the critic training parameters in our algorithm. Through the convergence analysis, it can be verified that all agents will converge to the same optimal model as the training time goes to infinity. To validate the implementation of our algorithm, a multiagent training framework is proposed to train each Universal Robot 5 (UR5) robot arm to reach the random target position. Finally, experiments are provided to demonstrate the effectiveness and feasibility of the proposed algorithm.

7 July 2025 in IEEE Transactions on Field Robotics
Verified
Journal Article
Closed access
MSMichal Staniaszek
Michal Staniaszek
TFTobit Flatscher
Tobit Flatscher
JRJoseph Rowell
Joseph Rowell
HNHanlin Niu
Hanlin Niu
WLWenxing Liu
Wenxing Liu
Nick HawesNick Hawes
Nick Hawes
This paper presents a technical overview of AutoInspect, a ROS-based software system for robust and extensible mission-level autonomy. Over the past three years we have deployed AutoInspect on multiple platforms in a variety of environments, from forests to fusion reactors, for durations ranging from hours to weeks. The AutoInspect system combines robust mapping and localisation with graph-based autonomous navigation, mission execution, and scheduling, into a complete autonomous inspection and monitoring system. Code for the graph-based autonomy component will be made available at ori.ox.ac.uk/projects/autoinspect. In this article we describe in detail our most commonly-used configuration of AutoInspect: a Boston Dynamics Spot fitted with a custom compute and sensing payload, called Frontier. To highlight the flexibility of the AutoInspect system, we also describe its deployment with different hardware and software configurations. We evaluate AutoInspect’s performance in two long-term deployments on Spot at the Culham Centre for Fusion Energy in Oxfordshire, UK. The first deployment took place at a robotics test facility, spanning 49 days, including 14 uninterrupted days of autonomous operation. The second deployment covered 35 days in the torus hall of the Joint European Torus (JET) fusion reactor, and achieved 15 days of uninterrupted operation. This was the first ever deployment of a fully autonomous mobile robot in a fusion facility.

4 July 2025 in IEEE Transactions on Vehicular Technology
Verified
Journal Article
Open access
WLWenxing Liu
Wenxing Liu
HNHanlin Niu
Hanlin Niu
KWKefan Wu
Kefan Wu
WPWei Pan
Wei Pan
ZJZe Ji
Ze Ji
RSRobert Skilton
Robert Skilton
When using deep reinforcement learning (DRL) to perform multi-robot exploration in unknown environments, the training model may produce actions that lead to unpredictable system behaviors due to the complexity and unpredictability of the surroundings. Therefore, ensuring safe exploration with DRL becomes critical. To tackle this issue, we propose a multi-agent framework that utilizes the formation scheme based on intermediate estimator compensation (IEC) to address the uncertainties introduced by DRL to ensure safe exploration. The convergence of the proposed scheme is verified via the Lyapunov method in the presence of tracking errors. An actor-critic-based DRL method is proposed for each mobile robot to deal with collision avoidance tasks. To enhance the efficiency of obtaining the DRL training model, a consensus-based training policy is introduced. The proposed safe learning framework successfully addresses uncertainties introduced by DRL while ensuring mapless exploration in both simulations and real-world experiments. The experimental video is available at: https://youtu.be/ce99FxKFFtY.

8 September 2023
Verified
Book Chapter
Closed access
PDF file
WLWenxing Liu
Wenxing Liu
HNHanlin Niu
Hanlin Niu
RSRobert Skilton
Robert Skilton
JCJoaquin Carrasco
Joaquin Carrasco

11 October 2023
Verified
Conference Paper
Closed access
PDF file
LTLei Tang
Lei Tang
YLYong Liu
Yong Liu
WLWenxing Liu
Wenxing Liu
XFXingsheng Feng
Xingsheng Feng
The point cloud data obtained by a 3D laser scanner is known for its high density, redundancy, and significant noise, which can result in a slow 3D reconstruction process and an imprecise reconstruction surface. To address these issues, a method for rapidly creating a 3D point cloud model based on 3D laser scanning was proposed. This approach involves removing outlier point clouds using a statistical filtering algorithm, downsampling the data using an improved octree voxel filtering algorithm, smoothing the point cloud using the moving least squares method and resampling it. Finally, the improved greedy projection triangulation algorithm is used for surface reconstruction. The experimental results show that the method proposed in this paper can improve the efficiency of model reconstruction, and the reconstructed model has smooth surface, accurate structure and improved precision.

17 September 2023 in arXiv
Verified
Preprint
Open access
PDF file
WLWenxing Liu
Wenxing Liu
HNHanlin Niu
Hanlin Niu
RSRobert Skilton
Robert Skilton
JCJoaquin Carrasco
Joaquin Carrasco
When transferring a Deep Reinforcement Learning model from simulation to the real world, the performance could be unsatisfactory since the simulation cannot imitate the real world well in many circumstances. This results in a long period of fine-tuning in the real world. This paper proposes a self-supervised vision-based DRL method that allows robots to pick and place objects effectively and efficiently when directly transferring a training model from simulation to the real world. A height-sensitive action policy is specially designed for the proposed method to deal with crowded and stacked objects in challenging environments. The training model with the proposed approach can be applied directly to a real suction task without any fine-tuning from the real world while maintaining a high suction success rate. It is also validated that our model can be deployed to suction novel objects in a real experiment with a suction success rate of 90\% without any real-world fine-tuning. The experimental video is available at: https://youtu.be/jSTC-EGsoFA.