Lei Zheng

Research Fellow · National University of Singapore

Safety-critical autonomy · robot safety · motion planning · safe learning-based control

About me Publications

🔥 News

Publications

Selected recent journal and conference papers

(2025). Occlusion-Aware Contingency Safety-Critical Planning for Autonomous Vehicles. In IEEE TCYB.

PDF Project Video DOI

(2025). Safe and Real-Time Consistent Planning for Autonomous Vehicles in Partially Observed Environments via Parallel Consensus Optimization. In IEEE TITS.

PDF Video DOI

(2024). Barrier-Enhanced Parallel Homotopic Trajectory Optimization for Safety-Critical Autonomous Driving. In IEEE TITS.

PDF Project Video DOI

(2024). Spatiotemporal Receding Horizon Control with Proactive Interaction Towards Autonomous Driving in Dense Traffic. In IEEE TIV.

PDF Video DOI

(2023). Real-Time Parallel Trajectory Optimization with Spatiotemporal Safety Constraints for Autonomous Driving in Congested Traffic. In ITSC.

PDF Video DOI

(2022). Safe Learning-Based Feedback Linearization Tracking Control for Nonlinear System with Event-Triggered Model Update. In IEEE RAL (Presentation at IEEE ICRA 2022).

PDF DOI

(2022). Safe learning-based gradient-free model predictive control based on cross-entropy method. In EAAI.

PDF DOI

(2021). Safe learning-based tracking control for quadrotors under wind disturbances. In ACC.

PDF Video DOI

See all publications

Projects

Research on safe autonomy, planning, and learning

*

Occlusion-Aware Contingency Safety-Critical Planning for Autonomous Vehicles

This research introduces a occlusion-aware contingency safety-critical planning approach for safety-critical autonomous vehicles. Leveraging reachability analysis for risk assessment, forward reachable sets of occluded phantom obstacles are computed. The occlusion-aware contingency planner then constructs multiple locally optimal trajectory branches (each tailored to different risk scenarios), and a shared consensus trunk is generated to ensure smooth transitions and motion consistency.

Safe and Real-Time Consistent Planning for Autonomous Vehicles in Partially Observed Environments via Parallel Consensus Optimization

This research introduces a consistent parallel trajectory optimization (CPTO) approach for real-time, consistent, and safe trajectory planning for autonomous driving in partially observed environments. The CPTO framework introduces a consensus safety barrier module, ensuring that each generated trajectory maintains a consistent and safe segment, even when faced with varying levels of obstacle detection accuracy. We validate our CPTO framework through extensive comparisons with state-of-the-art baselines across multiple driving tasks in partially observable environments. Our results demonstrate improved safety and consistency using both synthetic and real-world traffic datasets.

Barrier-Enhanced Homotopic Parallel Trajectory Optimization for Safety-Critical Autonomous Driving

Enforcing safety while preventing overly conservative behaviors is essential for autonomous vehicles to achieve high task performance. In this project, we propose a barrier-enhanced homotopic parallel trajectory optimization approach with over-relaxed alternating direction method of multipliers for real-time integrated decision-making and planning in cluttered driving environments. Through a series of experiments, the proposed development demonstrates improved task accuracy, stability, and consistency in various traffic scenarios using synthetic and real-world traffic datasets.

Multi-Modal Spatiotemporal Receding Horizon Planning for Autonomous Driving

This project presents a cutting-edge approach for safe and efficient autonomous driving in dense traffic scenarios. Our proposed Spatiotemporal Receding Horizon Control (ST-RHC) scheme generates dynamically feasible and energy-efficient trajectories in real-time, enabling vehicles to accurately perform complex driving tasks. The algorithm employs receding horizon optimization and iterative parallel methods to design a trajectory tree that optimizes planning and ensures proactive interaction to avoid accidents. We have implemented our algorithms on an autonomous car, successfully achieving vehicle following, lane changing, overtaking, and cruise driving in dense traffic flow simulations based on ROS2.

High-speed flight

Once received return control signal, the drone must generate smooth trajectories in real-time to avoid collision and be close to the reference spraying path to realize high-performance precision spraying in precision farming.

multi-agent

Multi-agent in precision farming.

Safe Learning

Designed efficient incremental Gaussian Processes accounting for airflow uncertainties. The wind disturbance caused by the external environment is estimated to improve flight safety and control stability in cluttered environments. Following that, the estimated wind disturbance is used to compensate for the associated control error.

Connected Cruise Control in Mixed Traffic Flow

For safety-critical vehicles in mixed traffic flow where most vehicles are human-driven, each autonomous vehicle keeps tracking its front vehicle at a desired constant speed while maintaining a safe following distance with it in normal situations. However, when a vehicle decelerates urgently in unexpected situations, the vehicle behind has to reduce its speed to avoid collision with the front vehicle. In these cases, there exists a conflict between safety and stable high-performance tracking. For safety-critical autonomous vehicles, safety must not be violated and the tracking errors should be kept as small as possible.

Safety-critical Agricultural Drone

To achieve high-speed autonomous flight of aerial vehicles and realize high-performance precision spraying in precision farming. Trajectories must be generated in real-time to avoid collision and be close to the reference spraying path. Because of the high navigation speed, short sensing range, and unknown environments, response time is extremely limited, making generating high-quality trajectories a significant challenge.

Experience

Research and industry

 
 
 
 
 

Research Fellow

National University of Singapore

Jan 2026 – Present Singapore
Department of Electrical and Computer Engineering, working with Prof. Armin Lederer on safe autonomy and learning-based control.
 
 
 
 
 

Visiting Scholar

Carnegie Mellon University — Robotics Institute

Feb 2025 – Aug 2025 Pittsburgh, PA, USA
Collaborated with Prof. Changliu Liu on fundamental robot safety and formal methods for autonomous systems.
 
 
 
 
 

Ph.D. in Robotics and Autonomous Systems

The Hong Kong University of Science and Technology

Sep 2022 – Oct 2025 Hong Kong
Research on safety-critical motion planning, contingency planning, and trajectory optimization for autonomous vehicles. Supervised by Prof. Jun Ma and Prof. Michael Yu Wang.
 
 
 
 
 

Senior Robotics Engineer

XAG

Jul 2021 – Jul 2022 Guangzhou, China
Developed real-time trajectory replanning and safe backup policies for agricultural UAVs; algorithms deployed in products used in 60+ countries and regions.

  • Motion control and trajectory optimization (modern C++)
  • Flight data analysis and planning/visualization tooling
  • Patents and technology transfer to field systems

Posts

News, talks, and paper announcements

1 min read
This paper presents a novel occlusion-aware Consistent Model Predictive Control (CMPC) strategy for mobile robots navigating in environments with occluded obstacles — situations where obstacles are hidden from the robot’s field of view but pose potential collision risks. The method strikes an effective balance between safety (avoiding collisions with occluded obstacles) and performance (maintaining smooth, consistent motion), enabling robots to navigate confidently in cluttered, partially observable environments.

1 min read
Human driver unpredictability forces a false choice between excessive caution and collision risk. We introduce event-triggered online learning that dynamically refines behavior estimates, adjusting safety margins based on observed actions. Our dual-trajectory framework with FRS-based barriers achieves zero collisions, 64% lower lateral jerk, and 14-20% higher speeds than conservative approaches—proving safety and efficiency can coexist without compromise.

1 min read
Autonomous vehicles navigating occluded intersections face unseen hazards. Our solution co-optimizes exploration and safety-fallback trajectories with a shared initial segment, enabling smooth transitions. Using reachability-based risk assessment and consensus ADMM, we achieve 30% faster traversal, non-stop motion, and 90% faster computation than baselines—validated on real robotic platforms.

2 min read
One key underlying factor to the safety concern stems from uncertainties in both internal system models and external environments. These uncertainties become critical in contingency scenarios, where potential risks cannot be predicted with certainty.

1 min read
Present a real-time fail-operational controller for autonomous driving in the presence of time-varying environmental disturbances. This controller is designed to guide autonomous vehicles back to a predefined safe state asymptotically, while upholding task efficiency.

1 min read
Present a real-time parallel trajectory optimization method for the autonomous vehicles to achieve high travel efficiency in dynamic and congested environments.
See all posts

Featured Publications

Real-time safe trajectory planning for autonomous vehicles in occluded environments is achieved through a contingency planner that …
PDF Project Video DOI

This framework introduces a consensus spatiotemporal safety barrier within a scenario tree structure to address potential hazards. …
PDF Video DOI

This research seamlessly integrates discrete decision-making maneuvers with continuous trajectory variables for safety-critical …
PDF Project Video DOI

This research proposes a computationally-efficient spatiotemporal receding horizon control (ST-RHC) scheme to generate a safe, …
PDF Video DOI

Present an integrated decision and trajectory planning approach for the autonomous vehicles to achieve high travel efficiency in …
PDF Video DOI

We propose a learning-based tracking control scheme based on a feedback linearization controller in which uncertain disturbances are …
PDF DOI

A safe and learning-based control framework for model predictive control is proposed to optimize nonlinear systems with a …
PDF DOI

Enforcing safety on precise trajectory tracking is critical for aerial robotics subject to wind disturbances. In this paper, we present …
PDF Video DOI

A learning-based MPFC control paradigm for nonlinear systems under uncertain disturbances, coupling a high-level model predictive path …
PDF DOI

A learning-based safety-stability-driven control algorithm is presented to guarantee the safety and tracking stability for nonlinear …
PDF DOI

Contact

Collaborations, speaking invitations, and opportunities

  • 4 Engineering Drive 3, National University of Singapore, Singapore 117583