Research Engineer (Robot Learning & Manipulation) - LYB7

Singapore Institute of Technology; Singapore; On-site; SGD 90,000 - 130,000; Full time

Job summary

Singapore Institute of Technology is seeking a Research Engineer (Robot Learning & Manipulation) to contribute to end-to-end robot learning pipelines for manipulation tasks. You will collect data, train policies, evaluate performance, and deploy on real robotic systems within an aviation-focused program.

Role requires hands-on experience with robot manipulators, ROS/ROS 2, and proficiency in PyTorch and Python.

Qualifications

• Bachelor’s degree or higher in Robotics, CS, Mechanical/Electrical Engineering, AI, or related field.
• Hands-on robotics research or development with physical robots.

Responsibilities

• Develop, implement and evaluate learning-based manipulation methods (imitation, RL) for contact-rich tasks.
• Own end-to-end robot learning pipeline: data collection, policy training, evaluation, deployment.
• Build real-world data collection pipelines via teleoperation or demonstrations.
• Train and deploy policies on robotic manipulators and mobile platforms; troubleshoot issues.

Skills

Robot learning

Imitation learning

Reinforcement learning

ROS/ROS 2

PyTorch

Python

C++

Simulation tools

Education

Bachelor's degree in Robotics or related field

Tools

ROS/ROS 2

MuJoCo

Gazebo

Job description

Research Engineer (Robot Learning & Manipulation) - LYB7 Job Description Research Engineer (Robot Learning & Manipulation) - LYB7 Posting Start Date: 07/09/2026 Schemes of Service: Research Division: Mechanical & Systems Engineering Employment Type: Fixed Term

Role Overview

As Singapore’s University of Applied Learning, SIT works closely with industry and research partners in pursuing applied research with real-world impact. Our research staff have opportunities to develop industry-relevant capabilities while contributing to multidisciplinary research projects.

The primary responsibility of this role is to support SIT’s research activities under a programme focused on robotics applications in the aviation industry. The Research Engineer will contribute to the end-to-end robot learning pipeline for manipulation tasks, including data collection, policy training, evaluation, and deployment on real robotic systems.

Key Responsibilities

• Develop, implement, and evaluate learning-based manipulation methods, including imitation learning, reinforcement learning, and related approaches for contact-rich robotic tasks.
• Own the end-to-end robot learning pipeline, from demonstration/data collection and dataset preparation to policy training, evaluation, real-robot deployment, and iterative improvement.
• Develop real-world data collection pipelines using teleoperation or other demonstration interfaces, including sensing, synchronization, logging, and data quality control.
• Train and deploy learned policies on robotic manipulators and mobile manipulation platforms, and troubleshoot performance issues across perception, control, learning, and hardware.
• Integrate robot hardware, sensors, compute platforms, perception, motion planning, and learning components into robust robotic systems.
• Design and conduct laboratory and field evaluations, including performance benchmarking, failure analysis, and reliability improvement.
• Work with researchers and industry partners to translate research into practical robotic capabilities, and contribute to technical documentation, publications, demonstrations, and project outcomes.
• Support technical evaluation and procurement of robotic platforms, sensors, compute infrastructure, and related equipment where required.
• To communicate and liaise with internal and external stakeholder to ensure project deliverable are met.
• Any other ad-hoc duties assigned by Supervisor.

Requirements

• Bachelor’s degree or higher in Robotics, Computer Science, Mechanical Engineering, Electrical Engineering, Control, Artificial Intelligence, or a related field.
• Relevant experience in robotics research, development, or industry, with substantial hands-on work on physical robotic systems.
• Demonstrated experience developing and deploying learning-based manipulation methods on real robots, rather than simulation-only experimentation.
• Practical experience with imitation learning, behavioural cloning, diffusion policies, reinforcement learning, or related robot learning approaches.
• Experience collecting and working with real-world robot demonstration data, including teleoperation or other human-in-the-loop methods.
• Strong hands-on experience with robotic manipulators, ROS / ROS 2, motion planning, calibration, trajectory execution, and real-robot debugging.
• Proficiency in PyTorch and Python, with working knowledge of C++ and good software engineering practices.
• Familiarity with robotics simulation environments such as Isaac Sim / Isaac Lab, MuJoCo, Gazebo, or equivalent.

Copyright © 2025 Singapore Institute of Technology. All Rights Reserved.