Robotics · AI · Vision
Pau Montagut Bofi
Robotics Intelligence student building at the intersection of embodied AI, computer vision, and real-world robots. Currently researching Vision-Language-Action models at TU Wien.
Robotics Intelligence student building at the intersection of embodied AI, computer vision, and real-world robots. Currently researching Vision-Language-Action models at TU Wien.
I'm a final-year Robotics Intelligence student at Universitat Jaume I, currently on a research internship at the V4R (Vision for Robotics) lab at TU Wien in Vienna, supervised by Tessa Pulli under Prof. Markus Vincze.
My work focuses on adapting Vision-Language-Action models to the Toyota HSR robot, bridging the gap between large-scale pretrained models and real-world robotic embodiments with non-standard action spaces.
Before Vienna, I co-founded the UJI Robotics Team, led technology at Adagio (an AI accessibility startup), and built systems ranging from ocean-cleaning robots to drone-based 3D reconstruction pipelines. I believe the best engineering feels effortless: simple on the surface, deeply considered underneath.
Research Intern, V4R Lab, TU Wien, Vienna
B.Sc. Robotic Intelligence, Universitat Jaume I, 2021–2026
Exchange, University of Bremen, 2024
1st Place ICRA 2026 VLA Competition · 1st Place ASTI Challenge 2023 · 2nd Place CEABOT 2022 · Premis Talent Award 2023
A selection of projects across robotics, AI, and computer vision, from research prototypes to competition-winning systems.
1st place out of 36 teams at the ICRA 2026 "From Data to Decisions" workshop competition. Fine-tuned SmolVLA and π0.5 on an 11-DoF Toyota HSR, achieving 60% success rate across 6 real-robot tasks. Proposed per-joint-group evaluation over standard MSE, with paper accepted as poster at ICRA 2026.
End-to-end pipeline: simulated drone captures multi-altitude orbital imagery, bypasses SfM with ground-truth poses, and trains a 3DGS model. Single-command automation via ROS2 + Gazebo + ArduPilot.
AI-powered platform that converts atypical speech into clear text and audio for meetings, classes, and video calls. Fine-tuned Whisper models for non-standard speech patterns. Leading technology strategy and product development.
Co-founded the university's competitive robotics team as undergraduate-only students. Competed in the European Rover Challenge against graduate and PhD-level teams from top universities, proving that ambition and teamwork can close the experience gap. Also achieved 1st place at ASTI Challenge 2023 and top positions at CEABOT. Multidisciplinary team of 15+ members.
Designed an autonomous robot to collect floating ocean waste using AI and computer vision. Won the best project award, attracting interest from municipalities and coastal cleaning companies.
Academic papers written during my studies, from drone reconstruction to embodied AI.
Whether it's about research, robotics, a collaboration, or just to say hi, I'd love to hear from you.
"Focus and simplicity. Simple can be harder than complex; you have to work hard to get your thinking clean to make it simple."Steve Jobs