Research Fellow (Modular Reconfigurable Mobile Robots)

National University of Singapore; Singapore; Vor Ort; SGD 70.000 - 110.000; Vollzeit

Zusammenfassung

National University of Singapore invites applications for a Research Fellow in the Modular Reconfigurable Mobile Robots (MR²) project. The role is based at the Kent Ridge Campus in Singapore within the Department of Electrical and Computer Engineering.

The successful candidate will design adaptive control, develop RL-based policies, and build scalable training environments using ROS 2, Gazebo and MuJoCo, while pursuing robust sim-to-real transfer across diverse robot configurations.

Qualifikationen

• PhD in Robotics, Control Engineering or related field.

Aufgaben

• Design adaptive and switching control algorithms for different robot configurations and real-time control.

Kenntnisse

Robot modelling

Control engineering

Reinforcement learning

Python

C

ROS 2

Gazebo

MuJoCo

Ausbildung

Ph.D. in Robotics, Control Engineering or related field

Tools

ROS 2

Gazebo

NVIDIA Isaac Sim/Isaac Lab

MuJoCo

Jobbeschreibung

Job Title: Research Fellow (Modular Reconfigurable Mobile Robots)

University-Level Unit: College of Design and Engineering

Faculty/Department-Level Unit: Electrical and Computer Engineering

Employee Category: Research Staff

Location_ONB: Kent Ridge Campus

Posting Start Date: 08/10/2026

Job Description

Research Fellow positions are open in the research group of Prof. Shuzhi Sam Ge at the Department of Electrical and Computer Engineering, National University of Singapore (NUS).

The position is aligned with Modular Reconfigurable Mobile Robots (MR²) project. The research will develop a reconfigurable control framework that combines automatic robot modelling, adaptive and switching control, and reinforcement learning-based self-adaptation. The goal is to enable MR² platforms to operate safely and efficiently across different robot configurations, driving modes, payloads, and environments.

The Successful Candidate Is Required To:

• Design adaptive and switching control algorithms for different configurations, mode transitions, trajectory tracking, and real-time robot control.
• Design and train deep reinforcement learning policies for robot motion control, mode selection, switching decisions, task adaptation, and autonomous skill acquisition.
• Develop scalable robot-training environments and simulation pipelines using ROS 2, Gazebo, NVIDIA Isaac Sim/Isaac Lab, and MuJoCo.
• Develop sim-to-real methods to improve policy generalization across robot configurations, payloads, terrains, and operating conditions.
• Integrate learning-based policies with model-based and safety-critical control methods, including constrained or safe reinforcement learning where appropriate.
• Implement and deploy control and learning algorithms in Python and C, integrate them with ROS 2 software and robot hardware, and validate performance on multiple MR² configurations in simulation and real-world experiments.

Qualifications

• Ph.D. degree in Robotics, Control Engineering, Electrical Engineering, Computer Science, Mechanical Engineering, Mechatronics, or a closely related discipline.
• Strong background in robot modelling and control, preferably including switched systems, adaptive or nonlinear control, and nonholonomic mobile robots.
• Hands-on experience with reinforcement learning for robotic systems.
• Strong programming skills in Python and C are required. Experience with Linux, Git, real-time implementation, and software integration is highly desirable.
• Familiar with ROS 2 and robot simulation/training platforms including Gazebo, NVIDIA Isaac Sim/Isaac Lab, and MuJoCo.
• Experience in sim-to-real transfer, domain randomization, hierarchical learning, transfer learning, model-based reinforcement learning, or safe reinforcement learning.
• A strong publication record, good analytical and problem-solving skills, and the ability to work independently and collaboratively in a multidisciplinary robotics team.