Daniele Calandriello

Politecnico di Milano, IIT@MIT, Massachusetts Institute of Technology

Papers

Total Citations

H-Index

About

Daniele Calandriello’s research lies at the intersection of humanoid robotics, imitation learning, and safe, adaptive motor control. His work focuses on enabling robots to acquire complex, multi-step skills from human demonstrations while respecting physical and environmental constraints. A key contribution is his development of **Constrained Dynamic Movement Primitives (DMPs)**, which combine the compact representation of movement primitives with reinforcement learning to ensure feasible skill execution on humanoid platforms—a foundational paper with 21 citations. Calandriello also advanced **minimal intervention control** for obstacle avoidance, allowing robots to learn from demonstrations while only intervening when necessary to prevent collisions (8 citations). His research on **sequencing multiple tasks with competing constraints** (7 citations) addresses the critical challenge of chaining single-stroke movements into coherent, constraint-aware behaviors. Beyond technical contributions, his early work on **Physically Interactive Robogames** (24 citations) established design guidelines for human-robot physical interaction, demonstrating a commitment to both theoretical rigor and practical, engaging applications. Calandriello’s work is essential reading for anyone interested in making humanoid robots more dexterous, safe, and capable of learning in the real world.