Doctoral Researcher in Humanoid Robotics, 3D Perception and Embodied AI

100%, Zurich

The Advanced Manufacturing Lab (am|z) at ETH Zurich, led by Prof. Markus Bambach, is seeking a highly motivated Doctoral Researcher to conduct research in humanoid robotics, 3D perception, and embodied artificial intelligence for autonomous manufacturing.

The position is embedded in an interdisciplinary research project aiming to enable humanoid robots to perceive, understand, and interact with complex manufacturing environments using novel RGB‑D sensing technologies.

We are looking for a candidate with a strong background in robotics, computer vision, or machine learning who is passionate about developing intelligent robotic systems and translating scientific concepts into working solutions.

The successful candidate will investigate novel approaches to robot perception, simulation‑to‑real transfer, and autonomous manipulation, contributing to the development of intelligent humanoid systems capable of performing complex manufacturing tasks.

The Advanced Manufacturing Lab at ETH Zurich investigates advanced robotic technologies for flexible, intelligent, and autonomous manufacturing.

This Doctoral Researcher is part of a collaborative research project focused on developing intelligent humanoid robotic systems using next‑generation, single‑chip RGB‑D sensing technologies.

The project aims to integrate novel depth‑sensing hardware with advanced perception algorithms, simulation environments, and learning‑based robotic control to enable autonomous and collaborative manufacturing operations.

The project combines fundamental research with experimental validation using physical humanoid robotic platforms and real manufacturing scenarios.

• 3D simulation and synthetic data generation:
  Development of realistic manufacturing environments and simulation pipelines for generating annotated RGB‑D data
• 3D semantic perception:
  Development of learning‑based methods for object detection, segmentation, and scene understanding using RGB‑D sensor information
• Sim‑to‑real transfer:
  Investigation of methods for transferring perception and manipulation capabilities from simulation to physical robotic systems
• Autonomous robotic manipulation:
  Integration of perception, planning, and control for robotic handling and assembly operations
• Collaborative humanoid robotics:
  Investigation of learning‑based coordination strategies, including multi‑agent reinforcement learning, for cooperative manufacturing tasks
• Language‑guided robotic actions:
  Exploration of approaches that enable robots to interpret natural‑language instructions and translate them into executable actions

• Conduct independent scientific research in humanoid robotics, 3D perception, and learning‑based robotic systems
• Develop and evaluate novel algorithms for RGB‑D‑based object detection, semantic segmentation, and 3D scene understanding
• Design and implement realistic robotic simulation environments and synthetic data generation pipelines for manufacturing applications
• Develop methods for improving the transfer of perception and manipulation capabilities from simulation to real‑world robotic systems
• Design and implement data acquisition, annotation, processing, and evaluation workflows for synthetic and real RGB‑D datasets
• Integrate perception algorithms with robotic manipulation, motion planning, and control frameworks
• Investigate learning‑based approaches to autonomous robotic decision‑making, with opportunities to explore multi‑agent coordination and language‑guided manipulation
• Implement, test, and validate developed methods on physical robotic and humanoid platforms
• Plan and conduct scientific…