Robot Control & Learning
Model-based control, reinforcement learning, learning from demonstration and hybrid architectures designed for reliable behaviour on real robotic systems.
Postdoctoral researcher · IIT
I design and test robotic systems at the intersection of control, machine learning, perception and physical hardware.
Profile
I am a postdoctoral researcher at the Industrial Robotics Facility (INBOT) of the Italian Institute of Technology (IIT), where I develop control methods for complex robotic systems. My current work focuses on tendon-driven continuum robots for the manipulation of soft and deformable objects.
My main research interest is the development of hybrid robotic control architectures that combine classical and model-based methods with reinforcement learning. The aim is to retain the structure, interpretability and reliability of established control techniques while using learning to improve adaptability and robustness against modelling errors, nonlinear dynamics and changing operating conditions.
I hold a PhD in Robotics and Intelligent Machines (DRIM), completed within the Italian doctoral programme of national interest coordinated by the University of Genoa, with research activities carried out at Università Politecnica delle Marche. My background combines industrial and mechanical engineering, robotics, artificial intelligence and hands-on experimental work.
I approach research as an engineer: models and algorithms must ultimately work on the real system. I am comfortable moving from mechanical design and assembly to control implementation, sensing, experimental validation and practical troubleshooting in the lab.
Expertise
My work connects algorithms, mechanics and experimentation. These are not separate interests, but parts of the same engineering process.
Model-based control, reinforcement learning, learning from demonstration and hybrid architectures designed for reliable behaviour on real robotic systems.
Tendon-driven continuum mechanisms, shape estimation, antagonistic actuation and manipulation strategies for soft and deformable objects.
Computer vision, visual feedback, markerless motion capture, 3D perception and sensor integration for observation and control.
Robot programming and integration for collaborative assembly, surface finishing, human–robot interaction and industrial automation.
Mechanical design, prototyping, actuation, electronics and system integration, with attention to how each subsystem affects the whole.
Test planning, data acquisition, system identification, validation and practical debugging under real laboratory constraints.
Software & engineering workflow
I work across Linux and Windows and select the environment according to the problem, moving between mechanical design, code, simulation, experimental deployment and technical documentation.
Code-based development for control, perception and machine learning, including ROS and ROS 2, MoveIt, OpenCV, simulation tools and experiment-specific software pipelines.
CAD, numerical modelling, control design and system analysis using Siemens NX, MATLAB and Simulink, together with specialised CAE, electronics and data-acquisition tools when required.
Version-controlled and collaborative workflows built around Git and GitHub, with experience maintaining codebases, experimental data and technical material across Linux and Windows environments.
Hands-on engineering
Mechanical work, assembly, wiring, calibration, robot setup, integration, testing and troubleshooting are a core part of how I develop and understand robotic systems.
Contact
I am always interested in technically meaningful discussions and collaborations.