NewTechOn-site
Robotics Software Engineer (Geometric Perception) – Working Student / Internship (m/f/d)
Job description
Your Role
driveblocks develops off-road ground autonomy that works under harsh and challenging conditions. We build systems designed for operation in 3D terrain, vegetation, and heavy dust and dirt, where 100% reliability is core to success, even in GNSS-denied environments.
Working with multiple OEM partners across agriculture, construction, and defense, our focus is to deliver value to vehicle operators and solve the toughest challenges when deploying Physical AI in real world applications. Every deployment helps us to gather the operational data required to build the next generation of reliable and safe autonomous driving systems.
From our office in Munich, we operate as a focused, highly ambitious team with deep expertise in autonomy, Physical AI, and embedded software engineering. We move quickly and set high standards. If you want to build autonomy that is deployed, tested, and operated in the field - we’d love to hear from you!
Working at driveblocks
- As a Working Student in the Geometric Perception team, you contribute to state-of-the-art perception algorithms deployed on autonomous vehicles operating in complex environments.
- Depending on your interests and the team’s priorities, your work may include:
- Developing and improving geometric and model-based perception algorithms, including sensor fusion, odometry, and environment understanding.
- Evaluating and benchmarking state-of-the-art open-source methods with a focus on robust classical, geometric and probabilistic approaches rather than deep learning.
- Designing automated validation and benchmarking tools to quantify software performance across different operating conditions, sensor setups, and vehicle platforms.
- Making our software more robust, efficient, and reliable for real-world use.
- Working with large-scale sensor datasets to analyze failure cases and drive performance improvements.
- This role offers the opportunity to build strong software engineering skills through hands-on work with tools and practices such as modern C++, Git, Docker, and agentic programming.
- Your work will directly contribute to the safe operation of autonomous heavy machinery in real-world applications.
- Bachelor’s degree in computer science, informatics, engineering or related fields.
- Practical experience in software development, e.g. through past work experiences, internships, Formula Student, etc.
- Interest in the fundamentals of robotics, geometric and probabilistic algorithms.
- Ideally, familiarity with Linux, C++, git, ROS2, containerized development processes, CI/CD frameworks.
- Availability of at least 15 hours per week.
driveblocks develops off-road ground autonomy that works under harsh and challenging conditions. We build systems designed for operation in 3D terrain, vegetation, and heavy dust and dirt, where 100% reliability is core to success, even in GNSS-denied environments.
Working with multiple OEM partners across agriculture, construction, and defense, our focus is to deliver value to vehicle operators and solve the toughest challenges when deploying Physical AI in real world applications. Every deployment helps us to gather the operational data required to build the next generation of reliable and safe autonomous driving systems.
From our office in Munich, we operate as a focused, highly ambitious team with deep expertise in autonomy, Physical AI, and embedded software engineering. We move quickly and set high standards. If you want to build autonomy that is deployed, tested, and operated in the field - we’d love to hear from you!
Working at driveblocks
- Ownership of your work – Take responsibility for the full value chain
- Competitive Compensation & Benefits
- Flexible working hours
- Upload your CV and additional documents directly through our platform
- Receive feedback on your application within a few days
- Show your motivation in a one-on-one conversation with us
Skills & technologies
Automated validationLarge-scale data analysisLinuxGitAgile developmentC++DockerSensor fusionOdometryCI/CD pipelinesModel-based perceptionProbabilistic algorithmsROS2