This thesis proposes a new method for robotic rehabilitation of the upper limb based on end-effector systems, with the aim of overcoming the inherent limi- tations of conventional Cartesian trajectory tracking approaches. Traditional TT-based systems reproduce the spatial trajectory recorded by the therapist at the end-effector level, but they do not guarantee that the patient executes the same joint behaviour. This discrepancy becomes particularly relevant when therapist and patient have different limb morphologies, since identical Carte- sian trajectories can correspond to markedly different joint configurations, with possible implications for safety, comfort, and therapeutic efficacy. To address this problem, this thesis introduces the Human Joint Trajectory Tracking (HJTT) algorithm, a method that explicitly operates in the human joint space. The core idea is to acquire the joint angles of the performing sub- ject (the therapist) throughout the demonstrated movement and use them as the fundamental input of the algorithm. The method is build on a paramet- ric kinematic model of the human arm and computes, at each time instant, the robot joint configuration that induces in the patient the same joint posture recorded from the therapist. The method was first validated in simulation to assess numerical stability, anthropometric consistency, and robustness to variations in limb geometry. It was then experimentally evaluated on two subjects with different anthropomet- rics, using Xsens inertial sensors to capture the therapist’s joint angles and a Franka Emika Panda robot to reproduce the adapted motion. The results show that HJTT achieves significantly improved joint tracking compared to stan- dard Cartesian tracking, demonstrating its potential to enhance personalization, safety, and therapeutic relevance in robot-assisted upper-limb rehabilitation. In addition, an existing graphical user interface was extended to support the new HJTT functionality, enabling clinicians to record, load, and adapt trajec- tories directly through the updated software.
Questa tesi propone un nuovo metodo per la riabilitazione robotica dell’arto superiore basato su sistemi end-effector, con l’obiettivo di superare i limiti in- trinseci dei tradizionali approcci di tracciamento della traiettoria cartesiana. I sistemi tradizionali basati sul TT riproducono la traiettoria spaziale registrata dal terapista a livello dell’end effector, ma non garantiscono che il paziente esegua lo stesso comportamento articolare. Questa discrepanza diventa par- ticolarmente rilevante quando il terapista e il paziente hanno morfologie degli arti diverse, poich´e traiettorie cartesiane identiche possono corrispondere a con- figurazioni articolari notevolmente diverse, con possibili implicazioni per la si- curezza, il comfort e l’efficacia terapeutica. Per affrontare questo problema, la presente tesi introduce l’algoritmo Hu- man Joint Trajectory Tracking (HJTT), un metodo che opera esplicitamente nello spazio articolare umano. L’idea centrale `e quella di acquisire gli angoli ar- ticolari del soggetto che esegue il movimento (il terapista) durante l’esecuzione di questo e di utilizzarli come input fondamentale dell’algoritmo. Il metodo si basa su un modello cinematico parametrico del braccio umano e calcola, in ogni istante di tempo, la configurazione articolare del robot che induce nel paziente la stessa postura articolare registrata dal terapista. Il metodo `e stato convalidato realizzando una parte in simulazione per va- lutare la stabilit`a numerica, la coerenza antropometrica e la robustezza alle variazioni nella geometria degli arti. `E stato poi valutato sperimentalmente su due soggetti con caratteristiche antropometriche diverse, utilizzando sensori in- erziali Xsens per catturare gli angoli articolari del terapista e un robot Franka Emika Panda per riprodurre il movimento adattato. I risultati mostrano che l’HJTT ottiene un tracciamento articolare significativamente migliorato rispetto al tracciamento cartesiano standard, dimostrando il suo potenziale per miglio- rare la personalizzazione, la sicurezza e la rilevanza terapeutica nella riabili- tazione robotica degli arti superiori. Inoltre, un’interfaccia grafica utente esistente `e stata estesa per support- are la nuova funzionalit`a HJTT, consentendo ai medici di registrare, caricare e adattare le traiettorie direttamente attraverso il software aggiornato.
Tracciamento della traiettoria guidato dalle articolazioni umane per una riabilitazione robotica avanzata basata sull'end-effector.
ROMEO, DANIELE
2024/2025
Abstract
This thesis proposes a new method for robotic rehabilitation of the upper limb based on end-effector systems, with the aim of overcoming the inherent limi- tations of conventional Cartesian trajectory tracking approaches. Traditional TT-based systems reproduce the spatial trajectory recorded by the therapist at the end-effector level, but they do not guarantee that the patient executes the same joint behaviour. This discrepancy becomes particularly relevant when therapist and patient have different limb morphologies, since identical Carte- sian trajectories can correspond to markedly different joint configurations, with possible implications for safety, comfort, and therapeutic efficacy. To address this problem, this thesis introduces the Human Joint Trajectory Tracking (HJTT) algorithm, a method that explicitly operates in the human joint space. The core idea is to acquire the joint angles of the performing sub- ject (the therapist) throughout the demonstrated movement and use them as the fundamental input of the algorithm. The method is build on a paramet- ric kinematic model of the human arm and computes, at each time instant, the robot joint configuration that induces in the patient the same joint posture recorded from the therapist. The method was first validated in simulation to assess numerical stability, anthropometric consistency, and robustness to variations in limb geometry. It was then experimentally evaluated on two subjects with different anthropomet- rics, using Xsens inertial sensors to capture the therapist’s joint angles and a Franka Emika Panda robot to reproduce the adapted motion. The results show that HJTT achieves significantly improved joint tracking compared to stan- dard Cartesian tracking, demonstrating its potential to enhance personalization, safety, and therapeutic relevance in robot-assisted upper-limb rehabilitation. In addition, an existing graphical user interface was extended to support the new HJTT functionality, enabling clinicians to record, load, and adapt trajec- tories directly through the updated software.| File | Dimensione | Formato | |
|---|---|---|---|
|
Human Joint Driven Trajectory Tracking for Enhanced End-Effector-Based Robotic Rehabilitation.pdf
accesso aperto
Dimensione
2.83 MB
Formato
Adobe PDF
|
2.83 MB | Adobe PDF | Visualizza/Apri |
È consentito all'utente scaricare e condividere i documenti disponibili a testo pieno in UNITESI UNIPV nel rispetto della licenza Creative Commons del tipo CC BY NC ND.
Per maggiori informazioni e per verifiche sull'eventuale disponibilità del file scrivere a: unitesi@unipv.it.
https://hdl.handle.net/20.500.14239/33681