Ph.D. Rodrigo Pérez Dattari

About

I am a researcher specialized in robotics, with a focus on imitation learning and control theory. My work aims to create robots that are easily programmable and adaptable to unstructured environments, ensuring safe operation while enabling non-experts to deploy them effectively. By integrating human feedback and inductive biases, I have developed solutions that enhance generalization, efficiency, and safety. I am eager to join a competitive team focused on advancing robotics in real-world applications.

Skills

Machine Learning for robotics: imitation/reinforcement learning, human-robot interactions, human-in-the-loop, safety guarantees (stability, collision avoidance), learning on Riemannian manifolds, and task planning via large language models.

Control: forward/inverse kinematics, impedance/velocity control, Operational space control, inverse dynamics control, model predictive control, visuomotor control.

Programming: ROS, python, deep learning (PyTorch, Tensorflow), git, C++, LaTeX, bash.

Project management: leading and contributing to collaborative research projects, supervising several undergraduate and graduate students.

Research & communication: authored multiple research papers and presented at international conferences (e.g., ICRA, CoRL, and ISER).

Languages: Spanish (native) and English (full professional proficiency).

Experience

2023-2024

Post-Doctoral Researcher @ Delft University of Technology

• Robotic imitation learning through dynamical-system-based priors, including non-Riemannian geometries, multistability, and limit cycles
• Large-scale task planning via large language models and graph-based knowledge representations

2019-2023

Ph.D. in Robotics @ Delft University of Technology

• Thesis: Generalizable Robotic Imitation Learning: Interactive Learning and Inductive Bias
• Interactive learning for data-efficient and human-friendly policy shaping
• Behavior priors based on control theory —MPC and dynamical systems theory— for increased data efficiency, reliability, and safety
• Combining several research projects into two proof-of-concept demonstrations: a greenhouse setup and a poultry processing setup

2016-2018

Development Engineer @ Papinotas (now Lirmi )

• Software and hardware projects for a school management app.

Education

2012-2019

Electrical Engineering — B.Sc. and M.Sc. @ University of Chile

• Focus on machine learning, robotics, control, telecommunications, and digital systems.
• Master thesis: Interactive Learning with Corrective Feedback for Continuous-Action Policies Based on Deep Neural Networks

Selected Projects

Dynamical systems for motion primitive learning

• Dynamical-system learning for robot control from demonstrations.
• Emphasis on introducing strong inductive biases in machine learning models.
• PyTorch-based implementation.

Interactive Imitation Learning

• Collection of works exploring robot policy learning methods from online human feedback (human-in-the-loop).
• Multiple deep learning frameworks and a comprehensive survey of the field.

KUKA iiwa control framework

• ROS-integrated framework featuring robot controllers tailored for research and educational applications.
• Codebase used as the backbone of multiple research projects.

Invited Talks and Lectures

Invited Talks

2024

• Generalizable Robotic Imitation Learning — Talk given at several German research institutes following the DAAD AInet Fellowship award . @ Bosch Center for Artificial Intelligence, @ Karlsruhe Institute of Technology (KIT), @ TU Darmstadt

Invited Lecturer

2024

• Course RO47019 Intelligent Control Systems @ Delft University of Technology — Lecture: Dynamical Systems for Imitation Learning

Interests

• Music (guitar, singing), hiking, chess, football

Publications

Journals

1. R. Pérez-Dattari, Z. Li, R. Babuška, J. Kober and C. Della Santina. “Leveraging LLMs, Graphs and Object Hierarchies for Task Planning in Large-Scale Environments”. (under review), 2024
2. R. Pérez-Dattari, C. Della Santina, and J. Kober. “PUMA: Deep metric imitation learning for stable motion primitives”. Advanced Intelligent Systems, 2024
3. R.Pérez-Dattari and J. Kober. “Stable motion primitives via imitation and contrastive learning”. IEEE Transactions on Robotics (T-RO), 2023
4. C. Celemin, R. Pérez-Dattari, E. Chisari, G. Franzese, L.de Souza Rosa, R. Prakash, Z. Ajanović, M. Ferraz, A. Valada, and J. Kober. “Interactive imitation learning in robotics: A survey”. Foundations and Trends® in Robotics, 2022
5. R. Pérez-Dattari, B. Brito, O. de Groot, J. Kober, and J. Alonso-Mora. “Visually-guided motion planning for autonomous driving from interactive demonstrations”. Engineering Applications of Artificial Intelligence, 2022
6. R. Pérez-Dattari, C. Celemin, G.Franzese, J. Ruiz-del Solar, and J. Kober. “Interactive learning of temporal features for control”. IEEE Robotics & Automation Magazine (RAM), 2020

Conferences

1. E.Drijver, R.Pérez-Dattari, J. Kober, C. Della Santina, and Z. Ajanović. “Robotic packaging optimization with reinforcement learning”. IEEE International Conference on Automation Science and Engineering (CASE), 2023
2. P. Valletta, R. Pérez-Dattari, and Jens Kober. “Imitation learning with inconsistent demonstrations through uncertainty-based data manipulation”. IEEE International Conference on Robotics and Automation (ICRA), 2021
3. R. Pérez-Dattari, C. Celemin, J. Ruiz-del Solar, and J. Kober. “Continuous control for high-dimensional state spaces: An interactive learning approach”. IEEE International Conference on Robotics and Automation (ICRA), 2019
4. R. Pérez-Dattari, C. Celemin, J. Ruiz-del Solar, and J. Kober. “Interactive learning with corrective feedback for policies based on deep neural networks”. International Symposium on Experimental Robotics (ISER), Springer, 2018
5. C. Celemin, R. Pérez-Dattari, J. Ruiz-del Solar, and M. Veloso. “Interactive machine learning applied to dribble a ball in soccer with biped robots”. RoboCup 2017: Robot World Cup XXI 11, Springer, 2018