(Senior) Research Engineer, Embodied AI & Robotics, IAIC

A*STAR RESEARCH ENTITIES; Singapore; On-site; SGD 90,000 - 150,000; Full time

Job summary

A*STAR RESEARCH ENTITIES is seeking a talented Senior Robotics R&D Engineer to advance manipulation and dexterity in robotic systems. You will tackle grasping, bimanual manipulation, and contact-rich tasks, shaping policies and controls for real hardware while collaborating with cross-functional teams.

You should bring a strong background in robotics, manipulation, and control, plus hands-on experience with ROS 2, deep learning, and real-time systems.

Qualifications

• Bachelor-level degree in a robotics, CS, EE, mechatronics or related field.
• Experience in developing manipulation, motion planning, or robot control systems.
• Experience deploying and debugging algorithms on physical robot hardware.
• Depth in grasp synthesis, manipulation planning, control, learning-based manipulation, and related perception tasks.

Responsibilities

• Develop and implement manipulation algorithms using learned, model-based, or hybrid approaches.
• Design and tune force, impedance, and hybrid force/position control for delicate objects.
• Develop perception for manipulation, including 6-DoF pose estimation and tactile sensing.
• Build simulation environments and drive sim-to-real transfer with system identification and domain randomization.
• Implement safety-aware control and motion planning for close human-robot interaction.
• Integrate hardware and software components into a ROS 2 system for real-world deployment.
• Define benchmarks to measure safety, performance and reliability.
• Collaborate with hardware, software and research teams to achieve project goals.
• Maintain documentation and mentor junior engineers.

Skills

Python

C++

ROS 2

Reinforcement learning

Imitation learning

6-DoF pose estimation

Tactile sensing

AI-assisted coding tools

Team collaboration

Real-time control

Education

Bachelor's degree in Robotics, Computer Science, Electrical Engineering, Mechatronics, or related field

Tools

Isaac Sim / Isaac Lab

MuJoCo

EtherCAT

Python

C++

Job description

Job Summary

We are seeking a talented and experienced (Senior) Robotics R&D Engineer to join our team, specialising in robot manipulation and dexterity. In this role, you will develop the grasping, bimanual manipulation and contact-rich dexterous in-hand manipulation capabilities that allow robots to handle diverse, fragile, tightly-toleranced and/or high-value parts — and to do so safely alongside people in shared workspaces. You will take manipulation tasks from concept to policies and controllers running on real hardware, collaborating with cross-functional teams, researching state-of-the-art methods, and validating system performance under real-world conditions. A strong foundation in robotics, manipulation and control, solid programming ability, and a rigorous experimental mindset are essential. Experience with dexterous or multi-fingered end-effectors, force/tactile sensing and ROS2 is an advantage, as is fluency with AI-assisted coding tools.

Roles and Responsibilities

You may be involved in a combination of the following tasks:

• Develop and implement manipulation algorithms for manipulation tasks, using learned (including but not limited to reinforcement learning, world model and other robot foundation models), model-based, or hybrid approaches as the task demands.
• Design and tune force, impedance, admittance, and hybrid force/position control to handle delicate and tightly-toleranced objects without damage.
• Develop perception for manipulation, including 6-DoF pose estimation, grasp affordance prediction, and tactile or multimodal sensing.
• Build simulation environments and drive sim-to-real transfer through system identification, domain randomization, and hardware validation.
• Implement safety-aware control and motion planning for close-proximity human-robot interaction.
• Perform integration of arms, grippers, dexterous hands, force-torque sensors, and tactile skins, integrating perception, learning, and control into a complete ROS 2 system for real-world deployment.
• Define and run benchmarks that measure system safety, performance, and reliability.
• Continuously research and implement state-of-the-art algorithms and methodologies to improve manipulation dexterity, robustness, and efficiency.
• Collaborate with cross-functional teams, including hardware engineers, software developers, and researchers in adjacent embodied AI areas, to achieve project goals.
• Maintain detailed documentation of code, algorithms, experiments, and project progress, and mentor junior engineers, facilitating knowledge sharing within the team.

Job Requirements

• Bachelor degree or higher in Robotics, Computer Science, Electrical Engineering, Mechatronics, or a related field.
• Demonstrated experience in developing manipulation, motion planning, or robot control systems.
• Demonstrated experience deploying and debugging algorithms on physical robot hardware.
• Demonstrated depth in at least two of the following areas, with working familiarity across the rest: Grasp synthesis, manipulation planning and control, Force, impedance, admittance, or compliance control, Learning-based manipulation, including imitation learning, reinforcement learning, and diffusion or vision-language-action policies, Tactile and multimodal perception for contact-rich tasks, Sim-to-real transfer and system identification.
• Experience with Isaac Sim / Isaac Lab, MuJoCo etc, and with real-time Linux, EtherCAT, or low-level motor control.
• Proficiency in Python and modern C++, with sound judgment on what belongs in a real-time control loop.
• Practical experience with ROS 2 in production or research systems.
• Knowledge of deep learning frameworks.
• Experience in software development, including version control, code review, testing, and debugging techniques on hardware.
• Strong communication skills to collaborate effectively with mechanical, controls, software, and process engineers, and to present technical findings.
• Strong commitment to solving difficult technical challenges in robot manipulation.
• Proficiency with AI coding assistants and agents as part of the development workflow is preferred.
• Publications at RSS, CoRL, ICRA, IROS, or equivalent, or significant open-source contributions is an added advantage.