My first goal in my Ph.D. is to develop a Simultaneous Localization and Mapping (SLAM) system to be applied in an indoor environment that has to be fully-autonomous once started. We think that for mobile robot companions being able just to navigate an environment is not enough. They need to understand where they are, have a memory of the past visited locations and react to changes in the environment (moved objects, inaccessible areas, new obstacles...). Moreover, they must do this in an autonomous way once powered-on.
To achieve this we are combining a Visual-SLAM system with a Model Predictive Control (MPC) mechanism to obtain area coverage and relocalization of the mobile platform (i.e. Active SLAM).
Afterward, this system will be further developed for example by adding human-robot and robot-environment interactions capabilities or cooperations between multiple agents.
Within the Robot Perception Group, I am also working with other members of the team in Multi-agent systems for aerial motion capture using perception (AirCap) and Deep Reinforcement Learning (AircapRL).
I've obtained my B.Sc. in Computer Science Eng. (2013) and M.Sc. in ICT for Internet and Multimedia Eng. (2019) at the University of Padua in Italy. Shortly after I started working as a research intern at MPI-IS in Tuebingen, Germany where now I am pursuing my doctoral degree as a member of the Robot Perception Group (RPG).
Autonomous Systems Active SLAM Reinforcement Learning Robotics
Autonomous MoCap systems, like AirCap, rely on robots with on-board cameras that can localize and navigate autonomously. More importantly, these robots must detect, track and follow the subject (human or animal) in real time. Thus, a key component of such a system is motion planning and control of multiple...
IEEE Robotics and Automation Letters, IEEE Robotics and Automation Letters, 5(4):6678 - 6685, IEEE, October 2020, Also accepted and presented in the 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). (article)
In this letter, we introduce a deep reinforcement learning (DRL) based multi-robot formation controller for the task of autonomous aerial human motion capture (MoCap). We focus on vision-based MoCap, where the objective is to estimate the trajectory of body pose, and shape of a single moving person using multiple micro aerial vehicles. State-of-the-art solutions to this problem are based on classical control methods, which depend on hand-crafted system, and observation models. Such models are difficult to derive, and generalize across different systems. Moreover, the non-linearities, and non-convexities of these models lead to sub-optimal controls. In our work, we formulate this problem as a sequential decision making task to achieve the vision-based motion capture objectives, and solve it using a deep neural network-based RL method. We leverage proximal policy optimization (PPO) to train a stochastic decentralized control policy for formation control. The neural network is trained in a parallelized setup in synthetic environments. We performed extensive simulation experiments to validate our approach. Finally, real-robot experiments demonstrate that our policies generalize to real world conditions.
Our goal is to understand the principles of Perception, Action and Learning in autonomous systems that successfully interact with complex environments and to use this understanding to design future systems