Current students


Section: Systems and Control

Major Research topic:
Increase human-robot awareness for a better cooperation in a collaborative environment

Collaborative robotics has an increasing trend in the automation industry and is expected to keep increasing in the following years. Up to now, companies use mostly collaborative robots (cobots) as normal robots but without fences since they are light and with ergonomic shapes so they are not able to create any harm to the human operator.
This is unluckily a very sub-optimal usage of cobots since their main advantages are not limited to safeties measures but are also endowed with high flexibility in the production process. This means that they can be combined with the dexterity of a human to achieve the most flexible system possible while at the same time they can be easily reprogrammed by a non-skilled operator.
This research aims at making the human-robot collaboration smoother and easier with a complete digitalization (digital twin) of all the components of the workspace: the robot, the human, the manipulated objects and the executed tasks.
With digital twin, it is meant the new concept introduced with Industry 4.0 where for each entity inside the workspace it is computed and constantly updated its digital representation which contains all the information that is known about it.
To have such representation an architecture that retrieves data from the working environment and update each digital twin has been devised. My research project is focused on improving this already existing architecture with new features and capabilities that can make the human-robot collaboration easier to achieve and, at the same time, give a straightforward interface to reprogram the robot to do a different task.
The architecture is mainly divided into two parts:

1) the first regards all the possible measurements and predictions it is possible to retrieve from the
working environment. For example, human intention prediction, pose estimation of the manipulated
objects, digital signals and human position occupation in the workspace. All of them are useful to
update each respective digital twin.
2) the second part models a generic assembly task through an SFC flowchart. This SFC combines all
the measurements cited above and, depending on the state of the SFC it is in, it sends to the robot
commands that enable it to do a particular task when certain conditions are satisfied. The conditions to
pass from a state of the SFC to the following one is deduced retrieving information that can be
obtained directly from the data of one digital twin or several.