1. BaSaMa & HiWi
    2.  

Development and Evaluation of Application Scenarios for a Quadruped Manipulator in NVIDIA Isaac Sim

Institute
Professur für autonome Fahrzeugsysteme (TUM-ED)
Type
Semester Thesis /
Content
experimental /  
Description

Motivation

Quadruped manipulators combine mobility and manipulation, enabling robots to operate in complex, unstructured environments such as retail stores, homes, and warehouses. While recent advances in simulation platforms like NVIDIA Isaac Sim allow efficient development and testing, there is still a lack of practical, application-driven scenarios that demonstrate their real-world potential.

Most existing work focuses on isolated capabilities (locomotion or manipulation), whereas real applications require integrated, task-oriented behaviors. This thesis aims to bridge this gap by developing realistic simulation scenarios that reflect real-world use cases and enable systematic evaluation and training of quadruped manipulators.

Goal

Develop an end-to-end simulation pipeline for application-driven evaluation of a quadruped manipulator in diverse real-world-inspired scenarios.

Specifically, the thesis aims to:

  • Design multiple application scenarios (e.g., retail assistance, home helper, warehouse logistics)
  • Build interactive environments in NVIDIA Isaac Sim
  • Integrate and configure a quadruped manipulator (sensors, control, tasks)
  • Enable task execution such as navigation, object picking, and human assistance
  • Evaluate performance across scenarios (success rate, efficiency, robustness)

Expected Deliverables

  • A set of application scenarios implemented in NVIDIA Isaac Sim
  • Integrated quadruped manipulator simulation setup (URDF/USD, sensors, control)
  • Task pipelines for navigation and manipulation
  • Evaluation report across different scenarios and tasks
  • Documented codebase (training, simulation, evaluation)

 

Required Skills

  • Good English or German proficiency
  • Strong programming skills (Python, ideally C++/ROS/ROS2)
  • Interest in robotics, simulation, and embodied AI
  • Basic knowledge of robotics (kinematics, perception, or control)
  • Experience with simulation tools or machine learning is a plus

Start

Work can begin immediately. If you are interested, please send a short motivation letter, CV, and transcript of records to: yuan_avs.gao@tum.de

Tags
AVS Gao
Possible start
sofort
Contact
Yuan Gao
yuan_avs.gaotum.de
Announcement