Research Engineer: Hierarchical Learning for ASVs

National University of Singapore; Singapore; On-site; SGD 100,000 - 160,000; Full time

Job summary

National University of Singapore is seeking a Research Engineer to work on hierarchical learning for a team of autonomous surface vehicles. The role involves developing learning-based controllers, implementing them in simulation and on hardware, and collaborating with partners on a water-based manipulation project.

The candidate should have a strong background in robotics, control, and computer science with emphasis on reinforcement learning, imitation learning, and generative motion modeling,

Qualifications

• Background in robotics, control engineering, or related field with a working foundation in classical control theory alongside modern learning-based methods.
• Excellent coding skills in Python with PyTorch, including reinforcement learning, imitation learning, and flow matching/diffusion-based generative models.
• Experience designing hierarchical control architectures that combine high-level RL policies with generative motion models and low-level tracking controllers.
• Familiarity with various conditioning architectures for generative sequence models.
• Hands-on experience with robotics simulation platforms (e.g. Isaac Lab, MuJoCo) and sim-to-real transfer for deploying learned policies on hardware.
• Good writing/spoken communication skills for writing research reports and presentations

Responsibilities

• Developing robust conventional and learning-based controllers for collaborative manipulation by a team of autonomous surface vehicles.
• Designing and implementing controllers in simulation and then on hardware with our partner

Skills

Python

PyTorch

Reinforcement learning

Imitation learning

Generative models

Flow matching

Diffusion models

Tools

Isaac Lab

MuJoCo

Job description

National University of Singapore is seeking a Research Engineer to work on hierarchical learning for a team of autonomous surface vehicles. The role involves developing learning-based controllers, implementing them in simulation and on hardware, and collaborating with partners on a water-based manipulation project.

The candidate should have a strong background in robotics, control, and computer science with emphasis on reinforcement learning, imitation learning, and generative motion modeling,