Research Assistant (SLAM and Perception for Robotic Assembly)

Interested applicants are invited to apply directly at the NUS Career Portal. Please note your application will only be processed if you apply via NUS Career Portal. NUS Career Portal link - https://careers.nus.edu.sg/job/Research-Assistant-%28SLAM-and-Perception-for-Robotic-Assembly%29/34487-en_GB/ We regret that only shortlisted candidates will be notified. Job Description The Department of Civil and Environmental Engineering at the National University of Singapore (NUS) invites applications for one (1) Research Assistant position in the Constructive Machines Lab. The lab builds robotic systems that autonomously build structures and studies the two-way relationship between how a structure is designed and how it is built. Its current focus is a team of mobile manipulators that transport and assemble scaffolding systems, temporarily support unstable partial assemblies, and progressively turn the structure-in-progress into navigable space for themselves. This position addresses the perception problem that follows from that premise: the robots operate in an environment they are themselves rebuilding, so the map cannot be assumed static. The successful candidate will develop simultaneous localisation and mapping for a world under construction, maintaining a consistent shared estimate of the structure's true as-built state across a team of robots, and closing the loop with the planning system so that perception is driven by what the plan actually needs to know. Work will begin in simulation and move onto the physical platforms as the lab's hardware comes online. The Research Assistant is expected to: • Develop and evaluate SLAM and state estimation methods for environments that change as the robots build, where newly placed structural elements must be incorporated into the map. • Perceive and track the partially built structure against the design model, including element-level pose estimation, deviation detection, and reporting the as-built state back to the planner. • Develop multi-robot state estimation, including map merging, co-localisation, and consistent reference frames across heterogeneous platforms. • Implement task-and-motion-planning-informed active perception, in which viewpoint and sensing decisions are selected according to the information the planner requires. • Benchmark methods in simulation and deploy them on the lab's mobile manipulation platforms in collaboration with the hardware team. Job Requirements • At least a Bachelor's degree in Computer Science, Electrical Engineering, Mechanical Engineering, Robotics, or a related field. • A Master's degree will be preferred. • Solid understanding of SLAM and state estimation fundamentals, including LiDAR-inertial and visual-inertial odometry, factor-graph optimisation, and loop closure. • Strong programming skills in C++ and/or Python, with working experience in ROS or ROS 2. • Experience with 3D reconstruction and scene representation, such as occupancy or TSDF mapping, point cloud registration, and mesh reconstruction. • Experience with learning-based perception, including object and structural element detection and 6-DoF pose estimation, with PyTorch or a comparable framework. • Familiarity with robotics simulation platforms such as Isaac Sim/Lab, MuJoCo, or Gazebo is preferred. • Experience deploying perception systems on real robots with LiDAR and camera sensing is highly preferred. • Ability to work both independently and collaboratively within a multidisciplinary research team. • Strong written and spoken communication skills. • Open to fixed-term contract.