Esther Luna Colombini

About

Esther Luna Colombini is a prominent Brazilian researcher whose work spans reinforcement learning, robotics, autonomous systems, and unmanned aerial vehicles. Based at the University of Campinas (UNICAMP), she has made significant contributions to the intersection of artificial intelligence and robotics, advancing both theoretical frameworks and practical implementations. Her most impactful contribution is a comprehensive survey on offline reinforcement learning (2023, 286 citations), which has rapidly become a foundational reference in the field, providing a structured taxonomy and identifying open research challenges. Her work on visual SLAM — notably LIFT-SLAM (123 citations) — demonstrates her expertise in deep learning-based perception for autonomous navigation. Early work applying proximal policy optimization to quadrotor control (52 citations) showcased her pioneering use of model-free RL for stabilizing inherently complex aerial systems. Colombini's research also extends to cognitive architectures for robotics, including attention-based models and consciousness-inspired frameworks for human-like robots, reflecting a deep interest in biologically motivated AI. Her involvement in the Brazilian Robotics Olympiad further highlights her commitment to science education and inspiring the next generation of roboticists. With a diverse and highly cited portfolio, she stands as a leading voice in intelligent autonomous systems research.

Research Focus

Key Achievements

Top Papers

1. 1
   A Survey on Offline Reinforcement Learning: Taxonomy, Review, and Open Problems

   286 citations · 2023
2. 2
   LIFT-SLAM: A deep-learning feature-based monocular visual SLAM method

   123 citations · 2021
3. 3
   Intelligent Control of a Quadrotor with Proximal Policy Optimization Reinforcement Learning

   52 citations · 2018
4. 4
   Drone Reconfigurable Architecture (DRA): a Multipurpose Modular Architecture for Unmanned Aerial Vehicles (UAVs)

   37 citations · 2020
5. 5
   A Survey on Offline Reinforcement Learning: Taxonomy, Review, and Open Problems

   31 citations · 2022
6. 6
   CONAIM: A Conscious Attention-Based Integrated Model for Human-Like Robots

   28 citations · 2016
7. 7
   Brazilian Robotics Olympiad

   19 citations · 2016
8. 8
   An Attentional Model for Autonomous Mobile Robots

   15 citations · 2016
9. 9
   Stable and fast model-free walk with arms movement for humanoid robots

   14 citations · 2017
10. 10
    A Novel Architecture for Multipurpose Reconfigurable Unmanned Aerial Vehicle (UAV): Concept, Design and Prototype Manufacturing

    10 citations · 2018

Key Collaborators

Contact & Links