Master-Thesis Projects

Summary

Auto-calibrazione mediante moto libero della posizione iniziale di un robot a cavi piano sottoattuato

Self-calibration by free motion of the initial position of an under-actuated cable-driven planar robot

This thesis proposes a new automatic calibration method that is applicable for under-actuated cable-driven parallel robots. The purpose of this work is to develop a method and a mathematical model that uses free motion as an exciting trajectory for the acquisition of calibration data. The key point of this method is to find a relationship between the unknown parameters to be calibrated (the lengths of the cables) and the parameters that could be measured by the proprioceptive sensors (the swivel pulley angles measured by the encoders). The equations involved are the geometrical-closure equations and the finite-difference velocity equations, solved using the least-squares algorithm.
Simulations are performed on a planar parallel robot driven by 2 cables for validation.

Status: in progress
Supervisors: Prof. M. Carricato, PhD. E. Idà

Movimentazione di elicotteri a bordo nave mediante l’utilizzo di veicoli autonomi ad ancoraggio magnetico

Movement of helicopters on board ship using autonomous vehicles with magnetic anchorage

This thesis proposes a new automatic calibration method that is applicable for under-actuated cable-driven parallel robots. The purpose of this work is to develop Study and design of a vehicle with magnetic anchorage to carry out handling operations of helicopters on board ship. The main phases of the project include the kinetostatic analysis of the helicopter, the dimensioning of the adhesion mechanism, and the conceptual design of the vehicle.

Status: in progress
Supervisors: Prof. M. Carricato, F. Malaguti
In collaboration with: L3HARRIS Calzoni

Ottimizzazione del prelievo fustelle di una macchina di confezionamento mediante algoritmi di apprendimento per rinforzo

Optimizing die picking in a packaging machine using reinforcement learning algorithms

This thesis addresses the challenge of introducing Artificial Intelligence into an industrial environment. The purpose of this work is to take advantage of the latest developments in the field of Reinforcement Learning to optimize the action of picking cardboard blanks from the buffer in a packaging machine for coffee capsules. In particular, the aim is to tackle the undesired behaviours under little variation of the features of the raw material (e.g., misalignment, curved due to humidity, etc.) using an adaptive controller. The new control system should be able to reproduce (and overcome) human understanding of which parameters of the electro-mechanical and pneumatic transmission must be changed in each situation to avoid or, at least, reduce erroneous picking cycles.

Status: in progress
Supervisors: Prof. M. Carricato, S. Comari, D. Baraccani
In collaboration with: IMA S.p.A. – R&I dept.

Human ​Upper-limb Workspace Computation by Interval Analysis for Ergonomics

The aim of the project is to define a kinematic model of the human arm, with particular interest on the solution of the inverse kinematics problem. After applying bio-mechanical constraints, it is possible to distinguish between feasible and non-feasible configurations of the human arm. This study is then extended to the use of interval analysis, in order to consider the variability of the parameters of a human arm among different people. In this context, classical Denavit-Hartenberg parameterization provides equations that are not suitable for interval analysis: for this reason, different methods to represent the pose of each link of the model are explored.

Status: in progress
Supervisors: Prof. M. Carricato, Dr. D. Daney, Prof. N. Rezzoug, R. Di Leva
In collaboration with: INRIA Bordeaux – Auctus Team

Progettazione e prototipazione del sistema di trasmissione per un Robot Parallelo a Cavi

Design and prototyping of a CDPR’s transmission system

This thesis objetive is to re-design and test a state-of-the-art transmission system developed for a Cable-Driven Parallel Robot (CDPR). By using a differential as a core part of the transmission it may be possible to remove one actuator of the robot. The first step toward an improvement of such innovative architecture is to analyse the current transmission, to find possible problems and design errors. Afterwards, one can proceed to re-design the transmission system, through the CAD software Creo Parametric, trying to fix any revealed issues and making the system more compact, light-weight, and precise. Moreover, the kinematic constraints required by the robot can be satisfied with different transmission systems structures. To this purpose, three transmissions system are built and compared in laboratory tests, highlighting the advantages and disadvantages of each one.

Status: in progress
Supervisors: Prof. M. Carricato, Dr. S. Caro, V. Mattioni
In collaboration with: LS2N ( Le Laboratoire des Sciences du Numérique de Nantes) – Centrale Nantes – RoMaS (Robots and Machines for Manufacturing, Society and Services) team