My research focuses primarily on: kinodynamic motion planning in dynamic and crowded environments, control of non linear dynamic systems, path and trajectory planning considering human motions predictions and socially normative behaviors. More specifically, in the last few years, I have been working with autonomous navigation systems in dynamic and crowded environments using sampling based motion planners. I find the latter to be very powerful frameworks where control theory, AI and machine learning based techniques intersect each other and help to solve difficult motion planning tasks.


Some contributions I have been working on (clik on the topic for further information): To diminish the gap between planning and control. For this topic I developed POSQ an ad-hoc steer function that improves the performances of sampling-based motion planner (RRT* and RRT). Integrate Motion Planning and Machine Learning techniques: to this end I introduced a fast way to compute the distance metric in the RRT planners by training a simple nonlinear parametric model with constant-time inference that is shown to predict distances accurately in terms of regression and ranking performance. Combine discrete path planners with continous single-query sampling-based planners by growing their trees in an informed and more focused way: this combination was shown to boost RRT planning performance and improve path quality for high dimensional nonholonomic wheeled mobile robots. Homotopy classes and path planning. Novel methods for real-time motion planning in dynamic environments.

Academic Activities

Member of IEEE RAS Technical Committee "Algorithms for Planning and Control of Robot Motion" Reviewer for IEEE RAS Transactions on Robotics Reviewer for IEEE International Conference on Intelligent Robots and Systems Reviewer for IEEE International Conference on Robotics and Automation.

Projects

ILIAD (Current)



The EU project ILIAD will develop robotic solutions for intralogistic services with key stakeholders from the food distribution sector. ILIAD pursues ambitious goals for complex cognitive systems in human environments beyond a specific use-case. We will overcome limitations in the state of the art in tracking and analysing humans; quantifying map quality and predicting future states depending on activity patterns inferred from long-term observations; planning of socially normative movements using learned human models; integration of task allocation, coordination and motion planning for heterogeneous robot fleets; and systematically studying human safety in mixed environments, providing a foundation for future safety standards. In this project, I am jointly responsible for the motion planning system development.



During the EU project SPENCER (www.spencer.eu) I was jointly responsible for the motion planning capabilities of a new socially aware passenger guidance robot. SPENCER's research covers key areas of interactive intelligent systems such as perception of people and groups of people in sensory data, normative human behavior learning and modeling, socially-aware mapping, and socially-aware task, motion and interaction planning in unstructured real-world environments and from mobile platforms.

In a final demonstration scenario we deployed a fully autonomous mobile robot for smart passengers flow management at the Amsterdam Schiphol Airport



During my Master Thesis at the Seconda Università degli Studi di Napoli, I was envolved in the Saphari EU FP7 Project, for the development of short range monitoring algorithms. The main goal of Saphari is to bring to fruition co-workers in real world applications using the new technologies of soft robotics that combine cognitive reaction and safe physical human-robot interaction.



During my Bachelor Thesis at the Seconda Università degli Studi di Napoli, I was envolved in the Dexmart EU FP7 Project, for the development of a network of microcontrollers for the digital linearization of sensors mounted on a robotics hand.

Embedded Systems in Critical Domains



After my Master Thesis I have been envolved in the project Embedded Systems in Critical Domains founded by SESM (a Finmeccanica company). The focus of the project includes the development of embedded systems in critical domains, in such framework I have been working on the topic of "Passive Coherent Localization": detection of human beings in cluttered indoor environment by exploiting machine learning and adaptive filtering techniques.


Past activities

2nd Prize in Altera Design Contest 2011 InnovateItaly. Member of the project “Real Time Control of an Anthropomorphic Robotic Arm using FPGA”. The project goal was the implementation of a control system for a 6DOF anthropomorphic arm using the FPGA DE1 Board produced by Altera

Finalist in Altera Design Contest 2010 InnovateItaly. Member of the project “Acquisition and processing system for real-time data generated by a tactile sensor”. The project goal was the implementation of a system for the digital linearization for a tactile sensor using the FPGA DE1 Board produced by Altera