Enhanced collaborative robotics via skill-based programming: the manipulation of linear deformable objects
Andrea Monguzzi
PHD Student
DEIB - Conference Room "E. Gatti" (Building 20)
June 13th, 2023
11.50 am
Contacts:
Simone Formentin
Research Line:
Control systems
PHD Student
DEIB - Conference Room "E. Gatti" (Building 20)
June 13th, 2023
11.50 am
Contacts:
Simone Formentin
Research Line:
Control systems
Abstract
On June 13th, 2023 at 11.50 am Andrea Monguzzi, PHD Student in Information Technology, will give a seminar on "Enhanced collaborative robotics via skill-based programming: the manipulation of linear deformable objects" in DEIB Conference Room.
The introduction of collaborative robots (cobots) brought significant changes in the paradigm of industrial robotics. Although implementing this technology in assembly tasks might be relatively straightforward, several important topics must be addressed to fully exploit its potentialities.
For example, the programming approach must be versatile: skill-based programming paradigms are the most promising since they allow cobots to be easily and quickly reprogrammed, even by users with limited expertise. This is due to the high-level description of the operations that are translated into competencies (the skills) that a given robot has and that can be concatenated to automatize complex tasks. The real challenge consists of building a rich set of skills. This research addresses one of the most challenging tasks in robotics: the manipulation of deformable linear objects (DLOs), such as cables, wires and ropes. Its ubiquity in industrial applications makes it a fundamental operation that is yet the bottleneck for several processes due to the high level of dexterity it requires. The manipulation of DLO is an under-constrained and non-trivial task as it must deal with problems in dynamics modelling, state representation and perception, while accounting for uncertainties. DLOs manipulation is hence a broad topic connecting different areas of robotics, like perception, simulation, control and mechanics.
The introduction of collaborative robots (cobots) brought significant changes in the paradigm of industrial robotics. Although implementing this technology in assembly tasks might be relatively straightforward, several important topics must be addressed to fully exploit its potentialities.
For example, the programming approach must be versatile: skill-based programming paradigms are the most promising since they allow cobots to be easily and quickly reprogrammed, even by users with limited expertise. This is due to the high-level description of the operations that are translated into competencies (the skills) that a given robot has and that can be concatenated to automatize complex tasks. The real challenge consists of building a rich set of skills. This research addresses one of the most challenging tasks in robotics: the manipulation of deformable linear objects (DLOs), such as cables, wires and ropes. Its ubiquity in industrial applications makes it a fundamental operation that is yet the bottleneck for several processes due to the high level of dexterity it requires. The manipulation of DLO is an under-constrained and non-trivial task as it must deal with problems in dynamics modelling, state representation and perception, while accounting for uncertainties. DLOs manipulation is hence a broad topic connecting different areas of robotics, like perception, simulation, control and mechanics.