Course paper / final thesis, Student placement

Analysis, Extraction, and Matching of Human Surface Processing Strategies to a Robotic Behavior.

Starting date

1 May 2023

Duration of contract

10 months

Type of employment


Surface processing involves tasks such as painting, grinding, and sanding. Usually, the robotic execution strategy that involves motions and forces is manually programmed and deployed to production systems that process then thousands of units per day. Although there are more than 30 years of efforts to automate these task efficiently, there is still a lack of automation found in the case of small lot sizes and in the craftmenship area. Considering that the task of manual grinding or sanding is cumbersome, physically demanding and subject to health risks due to material residuals and dust, it is of major interest to automate such scenarios as well.

In surface processing, a so-called coverage path is used to plan the motion with the tool's processing imprint on the whole surface of the workpiece. The goal is a good coverage of the surface while reducing the overlap between adjacent paths. On the technical side, there are often assumptions made about a good strategy, such as meandering pathways. However, a human decides about the path based on the made experiences in the field. Therefore, a robot could possibly benefit from the human strategy to improve the coverage path on the workpiece surface.

The goal of this project is to collect data from human workers, which are ideally craftsmen, that process a workpiece surface by sanding it. The collected data shall investigated with respect to reoccurring features (analysis). Next, a number of common strategies shall be extracted by clustering the human demonstrations (extraction). This allows to train a model, which is able to predict the strategy from user input. This information can be used to match the human strategy onto a predefined robot behavior (matching) that is used in the task execution.

After achievment of the first two project milestones, which are the collection of human demonstration data and a first data cluster analysis, this position could be continued as a master thesis.


  • Literature survey
  • Collection of human demonstration data from sanding tasks
  • Data cluster analysis
  • Extraction and modelling of human expert strategies
  • Development of a matching algorithm between user demonstration and robot behavior
  • Experiments with a robotic system
  • Documentation and presentation


Your qualifications:

  • Knowledge about foundations of robotics
  • Experience in data analysis and machine learning, especially clustering and classification
  • Excellent Python skills, knowledge about C++
  • Experience wit Linux

Your benefits:

Look forward to a fulfilling job with an employer who appreciates your commitment and supports your personal and professional development. Our unique infrastructure offers you a working environment in which you have unparalleled scope to develop your creative ideas and accomplish your professional objectives. Our human resources policy places great value on a healthy family and work-life-balance as well as equal opportunities for persons of all genders (f/m/x). Individuals with disabilities will be given preferential consideration in the event their qualifications are equivalent to those of other candidates.

  • Apply online now
  • You can send this job advertisement via e-mail and complete your application on a personal computer or laptop.

    We need your digital application documents (PDF). The document upload function is not supported by all mobile devices. Please complete your application on a PC/laptop.

    Complete application on PC

Technical contact

Thomas Eiband
Institute of Robotics and Mechatronics

Phone: +49 8153 28-3341

Send message

Vacancy 79723

HR department Oberpfaffenhofen

Send message

DLR site Oberpfaffenhofen

To location

DLR-Institute of Robotics and Mechatronics

To institute