This thesis presents a simulation-supervised optimization framework for the optimal de- ployment of static robotic contact-based operations, such as screwing and drilling. The proposed approach aims at overcoming experience-based setup procedures by introducing an automated method that identifies the optimal workcell design. This framework focuses on static tasks, in which the dynamic components can be considered negligible. Because of the complexity of the problem, the proposed framework is formulated as a bi-level optimization framework, adopting a leader–follower paradigm. In particular, the lower level problem requires solving the inverse kinematics of the robot at each target location. The algorithm is necessary for an efficient exploration of the search space of the joint configurations. Instead, the leader problem is responsible for identifying the optimal configuration of the workcell design parameters. The goal of the proposed procedure is to organize the resources associated with the workcell layout, with the clear intent of finding a torque- efficient solution with a particular focus on the manipulability metric. The pipeline was designed to be easily reconfigurable, allowing its application to a wide range of static tasks performed by different robotic manipulators. This framework was tested in a screw-tightening scenario and in a drilling one. In both analysis, the proposed procedure was able to find optimal configurations.
La presente tesi propone un framework di ottimizzazione supervisionato tramite simulazione per il deployment ottimale di celle di lavoro robotiche dedicate a operazioni di contatto statico, come avvitatura o foratura. L’approccio proposto mira a sostituire le procedure basate sull’esperienza dell’operatore, introducendo un metodo automatico che identifica il design ottimale dell’area di lavoro. Questo framework si focalizza su operazioni statiche, nelle quali le componenti dinamiche del manipolatore possono essere trascurate. Data la complessità del problema, la struttura risolutiva presenta un’organizzazione a due livelli; in particolare, viene adottato un paradigma leader–follower. Il livello più basso si occupa della risoluzione del problema dell’inversione cinematica del robot per ogni posizione target. L’algoritmo deve esplorare efficientemente lo spazio di ricerca dei giunti del manipolatore. Il problema leader è responsabile dell’identificazione delle config- urazioni ottimali relative al design dell’area di lavoro. L’obiettivo del framework è quello di disporre le risorse associate al layout della cella in modo ottimale, trovando soluzioni efficienti dal punto di vista delle coppie ai giunti richieste al robot per eseguire la relativa operazione, con un focus sull’indice di manipolabilità del robot. La struttura è stata sviluppata con il fine di essere facilmente riconfigurabile, permettendo quindi di essere impiegata su un ampio range di operazioni statiche e su diversi robot. Il framework è stato testato su due scenari, il primo riguarda un’operazione di avvitatura, mentre il secondo di foratura. In entrambi i casi, la procedura proposta è stata in grado di trovare soluzioni ottime.
A simulation-supervised optimization framework for the optimal deployment of static robotic contact-based operations
Casciani, Alessandro
2024/2025
Abstract
This thesis presents a simulation-supervised optimization framework for the optimal de- ployment of static robotic contact-based operations, such as screwing and drilling. The proposed approach aims at overcoming experience-based setup procedures by introducing an automated method that identifies the optimal workcell design. This framework focuses on static tasks, in which the dynamic components can be considered negligible. Because of the complexity of the problem, the proposed framework is formulated as a bi-level optimization framework, adopting a leader–follower paradigm. In particular, the lower level problem requires solving the inverse kinematics of the robot at each target location. The algorithm is necessary for an efficient exploration of the search space of the joint configurations. Instead, the leader problem is responsible for identifying the optimal configuration of the workcell design parameters. The goal of the proposed procedure is to organize the resources associated with the workcell layout, with the clear intent of finding a torque- efficient solution with a particular focus on the manipulability metric. The pipeline was designed to be easily reconfigurable, allowing its application to a wide range of static tasks performed by different robotic manipulators. This framework was tested in a screw-tightening scenario and in a drilling one. In both analysis, the proposed procedure was able to find optimal configurations.| File | Dimensione | Formato | |
|---|---|---|---|
|
2025_12_Casciani_Thesis.pdf
accessibile in internet per tutti a partire dal 17/11/2026
Descrizione: Thesis
Dimensione
26.88 MB
Formato
Adobe PDF
|
26.88 MB | Adobe PDF | Visualizza/Apri |
|
2025_12_Casciani_Executive_Summary.pdf
accessibile in internet per tutti a partire dal 17/11/2026
Descrizione: Executive Summary
Dimensione
1.99 MB
Formato
Adobe PDF
|
1.99 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/247118