Research Areas and Collaborations

Motion Planning For Human-Robot Cooperation

Step-changes in safety technologies have opened robotic cells to human workers in real industrial scenarios. However, the lack of methodologies for productive and effective motion planning and scheduling of human-robot cooperative (HRC) tasks still limits the spread of HRC systems. Standard methods fail due to the high variability of the robot execution time, which is caused by the need to modify the robot’s motion to ensure continuous human safety. In this context, my research ambitions are multiple. First, to deploy better and innovative motion planners both for robot arms and mobile manipulators, able to exploit the human behavior models to increase the ergonomics and safety togheter with the robot system productivity. Second, to better integrate the motion planning with the task and motion planning to optimize not just locally the execution of one task, but the entire process.

Physical Human-Robot Interaction

The industry needs controllers that allow smooth and natural physical Human-Robot Interaction (pHRI) to make production scenarios more flexible and user-friendly. Within this context, my research have been focused over the years on multiple methodologies. Starting from the development of adaptive impedance controllers, I then moved to Role Arbitration methodologies, which are the mechanism that assigns the role of the leader to either the human or the robot. My team and I investigated deeply the adoption of the Game-Theory (GT) to model pHRI, and specifically, Cooperative Game Theory (CGT) and Non-Cooperative Game Theory (NCGT). In such a context, we also investigated how to model the human behavior and intention when interacting with the robot.

Human-Robot Safe Interaction

Activities Development of control algorithms for collision avoidance and online rescheduling based on the estimated position of people within the robot’s operating space. Development of methodologies for functional safety certification of non-certified sensor networks.

Robotized Processes And Control

Investigation on adopting industrial robots for additive and traditional technologies and processes. Laser-based direct metal deposition (LBDMD) is a promising additive manufacturing technology well suited for producing complex metal structures, low-volume manufacturing, and high-value component repair or modification. It finds broad application in the automotive, biomedical, and aerospace industries. The research focuses on optimizing path planning when redundant axes (robot + external axes) are present, and the operations are only partially constrained in Cartesian space. Furthermore, the integrated control of the robot’s motion and the process may dramatically increase the results in additive and subtractive technologies. The research also investigates methodologies on virtual sensors that exploit the knowledge of hindsight robots about the process.

Elasto-Dynamics Modelling And Control

The attempt to use industrial robots for technological and interaction tasks, i.e., robotic machining and automatic assembling, implies, on the one hand, the knowledge of the interaction force, on the other hand, the reduction of physical sensors. The research aims to define the best models with lumped parameters (both rigid and elastodynamic models) to predict the behavior of a manipulator’s cups during the execution of generic trajectories. Integration of advanced models in robust/optimal adaptive controls. Furthermore, this work aims to develop a virtual force sensor to estimate the interaction force.

Rehabilitation Robotics

The increasing interest from the medical rehabilitation world towards robotic technologies to recover neuromotor functions has led to the development of the Rehabilitation Multi-Sensory Room (RehaMSR), a multisensory robotic platform to be used as an aid to neuromotor rehabilitation for the execution of rehabilitation exercises. This platform consists of hardware and software components coordinated by a central control system, interacting with the patient and adapting to it.

Robot Kinematics Analysis And Synthesis

Research in the optimum design has taken different directions. One of those was to define the kinematic or dynamic parameters that determine the characteristics of the manipulator to justify the best design. In most of the studies underway, the possible solutions are restricted to one feasible region in which all geometrical and dynamic restrictions and the drives’ power input must be met. The research aims to investigate solutions through an energy approach where redundant actuation and elastic elements in the chain achieve the optimum design in a multiple-link system.