DIVA
2020-2022, Germany
Soft Skills:
Project Management, Leadership, Cross-functional Team Collaboration, Problem-Solving, Effective Communication
Tools:
Microsoft Teams, Microsoft Office, GitLab, Trello
Technical Skills:
CAD Modeling, Prototyping, Systems Design, Robotics Engineering, R&D
Tools:
Fusion 360, Inventor, GitLab, ROS (Robot Operating System), ANSYS
Description
DIVA, an acronym for Intuitive Design in Contrast to the V-model and its Analysis, embarked on an investigative journey into the realm of robotic design methodologies. This project was a comparative study between the traditional Bottom-Up, often termed Expert/Intuitive Design, and the more structured Top-Down approach mirrored after the V-Model commonly utilized in the automotive and aerospace sectors. The Bottom-Up approach typically relies on the designer's intuition and experience, leading to a sequential and iterative development process. In contrast, the Top-Down strategy, as encapsulated by the V-Model, advocates for a systematic and holistic view, starting from high-level requirements down to the minutiae of design specifications. DIVA's core objective was to scrutinize these methodologies, assessing their efficacy in the complex task of humanoid robot arm design. The project sought to uncover the intricacies and potential advantages of a methodical design process that could potentially yield non-intuitive solutions within the robot's design space, thereby challenging the conventional wisdom of robotic design.
By adopting the V-Model, the project was able to identify a solution space that provided a range of permissible design variables, thereby facilitating the physical realization of the robot arm with greater flexibility and robustness. This method allowed for the decomposition of high-level requirements into sub-system-level specifications with tolerances, effectively addressing the "chicken-or-egg" conundrum often encountered in design sequences.
Results
The solution space approach enabled the parallel development of different subsystems, significantly reducing the overall development time. One of the most groundbreaking outcomes of DIVA was the ability to decouple the robot's morphology from its control systems, which traditionally are tightly interwoven. This decoupling granted designers the freedom to explore a wider array of mechanical configurations without being constrained by control parameters. The project's approach also provided insights into the trade-offs within the design space, allowing for a more informed selection of control variables that could be adjusted to expand the ranges of other design variables. In essence, DIVA's exploration into the V-Model approach yielded a more robust, flexible, and efficient design process for humanoid robots, potentially setting a new standard in robotic systems design.
Publications:
Sathuluri, Akhil, Anand Vazhapilli Sureshbabu, and Markus Zimmermann. "Robust co-design of robots via cascaded optimisation." In 2023 IEEE International Conference on Robotics and Automation (ICRA), pp. 11280-11286. IEEE, 2023.
Sathuluri, Akhil, Anand Vazhapilli Sureshbabu, and Markus Zimmermann. "A systems design approach for the co-design of a humanoid robot arm." arXiv preprint arXiv:2212.14256 (2022).
Krischer, Lukas, Anand Vazhapilli Sureshbabu, and Markus Zimmermann. "Modular topology optimization of a humanoid arm." In 2020 3rd International Conference on Control and Robots (ICCR), pp. 65-72. IEEE, 2020.
