Bernardo Aceituno-Cabezas

Massachusetts Institute of Technology, Universidad Simón Bolívar

Papers

Total Citations

H-Index

About

Bernardo Aceituno-Cabezas is a leading researcher in robot manipulation and locomotion, whose work bridges the gap between discrete decision-making and continuous optimization for complex robotic systems. His primary research areas include contact-rich manipulation, multilegged locomotion planning, and certified grasping. Aceituno-Cabezas is best known for pioneering optimization frameworks that treat contact as a decision variable rather than a fixed constraint. His most influential work, "A Global Quasi-Dynamic Model for Contact-Trajectory Optimization in Manipulation" (30 citations), introduces a global optimization model that determines how a robot should make and break contact to achieve a desired object trajectory, reasoning over simplified geometric environments. Earlier, he developed a mixed-integer convex optimization framework for robust multilegged locomotion over challenging terrain (14 citations), which overcame the limitations of fixed gait sequences and ZMP-based stability. This work was generalized in a subsequent paper (6 citations) to create a continuous optimization approach for footstep planning on uneven terrain. His recent work on "Certified Grasping" (4 citations) provides formal guarantees for grasp success. Aceituno-Cabezas’s contributions are foundational for enabling robots to autonomously plan and execute complex physical interactions in unstructured environments.