Master's Thesis - Localization of Mobile Tracked Robots in Open Areas Using Remote Visual Features

• Review of the state of the art regarding localization in feature-poor outdoor environments (specifically Visual SLAM) and place recognition methods, e.g., using foundation models and vision-language models
• Selection and evaluation of suitable algorithms based on robustness, latency, and deployability on the robot
• Design of a system architecture that integrates the bridging concept into the localization stack, including the detection of LiDAR degeneration and mode transitions
• Design and prototypical implementation of a mechanism to capture distant visual features during phases of reliable localization
• Experimental validation in simulation or on a real robot

What you contribute, * Hands-on work with our robots in our own test environments in Stuttgart

• Opportunity to take on responsibility and the freedom to implement your own ideas
• Collaboration with top students in your field

We value and promote the diversity of our employees’ skills and therefore welcome all applications - regardless of age, gender, nationality, ethnic and social origin, religion, ideology, disability, sexual orientation and identity. Severely disabled persons are given preference in the event of equal suitability. Our tasks are diverse and adaptable - for applicants with disabilities, we work together to find solutions that best promote their abilities. Remuneration according to the general works agreement for employing assistant staff.

With its focus on developing key technologies that are vital for the future and enabling the commercial utilization of this work by business and industry, Fraunhofer plays a central role in the innovation process. As a pioneer and catalyst for groundbreaking developments and scientific excellence, Fraunhofer helps shape society now and in the future.

• Currently enrolled at a German college or university
• Proficiency in C++/Python
• Experience with ROS, image processing, point cloud processing, or machine learning is an advantage
• Analytical mindset and experience in algorithm development
• Enthusiasm for mobile robotics
• Excellent German or English language skills

The Fraunhofer-Gesellschaft (www.fraunhofer.com) is one of the world’s leading organizations for application-oriented research. 74 institutes develop pioneering technologies for our economy and society - more precisely: 30 000 people from technology, science, administration and IT. They know: Anyone who comes to Fraunhofer wants to and can make a difference. For themselves, for us and for the markets of today and tomorrow.

Call for applications for disciplines such as: electrical engineering, computer science, control systems, or related fields.

In the “Navigation mobile Robots” research group, we develop autonomous outdoor robots such as tracked platforms for applications in exploration, agriculture, and forestry, capable of independently traversing vast terrains. The group focuses on developing both the robots themselves and the autonomy stack that powers them. Outdoor robots operating in large open areas (e.g., airports, agricultural sites, or remote coastal regions) face a fundamental limitation: LiDAR-based localization degrades in open spaces due to a lack of geometric features, while GNSS/RTK signals are not always available or reliable. Consequently, missions must either be interrupted or continue with uncontrolled, accumulating drift.

The objective of this thesis is to develop and evaluate a “bridging” localization method that enables the robot to continue its mission even in areas where LiDAR localization degrades. The approach combines multiple sources of information:

• Distant visual features or those visible on the horizon, potentially captured and referenced automatically during the initial phase of the mission when LiDAR localization was reliable
• Semantic image features derived from foundation models, where applicable, to ensure robust (context-aware) recognition of these visual features
• The robot’s proprioceptive sensors: IMU, odometry/encoders
• The robot’s planned movement patterns, serving as a trajectory prior