Vedant Dave
Postdoctoral Researcher, TU Wien
Robot learning and Tactile manipulation
I am a postdoctoral researcher at the Autonomous Systems Lab, TU Wien, with Prof. Dongheui Lee, where my research focuses on tactile manipulation, residual world models, test-time adaptation, vision-language-action models, and diffusion policies, with a broader interest in robust and generalizable robot learning.
Previously, I was a postdoctoral researcher at the Chair of Cyber-Physical-Systems, Montanuniversität Leoben, where I conducted research on robot learning and supervised Ph.D. and Master's students. I received my Ph.D. in 2026 from Montanuniversität Leoben under the supervision of Prof. Elmar Rückert, with a dissertation on robust robot learning through task-structured representations. During my doctoral studies, I was also a research intern at the Okinawa Institute of Science and Technology (OIST) in 2025, working with Prof. Makoto Yamada.
Before my doctoral studies, I earned an M.Sc. in Automation and Robotics from TU Dortmund, completing my Master's thesis at Mercedes-Benz AG. I also worked as a research intern at the Bosch Center for Artificial Intelligence with Leonel Rozo.
Refresh the page to see Vedant in a different setting.
Inspired by Homanga.
Papers
Papers under review are not listed.
Collaborations
With industry
-
Stahl- und Walzwerk Marienhütte GmbH
Predicted the yield strength of steel from production process data, designed the neural networks and optimisation, and traced measurement inaccuracies through data analysis.
With other research groups
-
Deep reinforcement learning for automated decision-making in wellbore construction
Deep RL for automating drilling decisions, aimed at better control and safety.
-
Robust representations during sensory loss
Multi-view disentanglement for diffusion-based policies, so they tolerate sensor noise and failures.
-
Physics-informed neural network for predicting Gibbs free energy
Coding and supervision. Combines a physics-informed neural network with the CALPHAD formalism to predict the Gibbs energy of alloys.
-
Green Facade
Coding and supervision. Predicts floor level from the distribution of trace elements in Vienna's urban aerosol.
Teaching
At Montanuniversität Leoben, 2022 to 2026.