Teleoperation systems offer a safe method for humans to interact with environments that were previously unreachable due to several safety reasons. It is characterized by a communication channel connecting the commending robot to the commended one. A Unilateral Teleoperation system is characterized by information being sent from the commending robot to the commend one, only. In a bilateral Teleoperation system, this exchange happens from both sides, where the commended robot also sends signals (usually sensor information about the remote environment) to the human operator. However, communication channels can introduce time delays, especially in scenarios where a large distance is present between both parties, leading to instabilities in the system. In this paper, we offer a continuation based on the results of \cite{1626803}. We test the proposed architecture without making assumptions on the nature of the time delay, on a more complex system consisting of a Panda 7 DoF robot \cite{Franka} as the remote robot and a Human operator equipped with a WEART TouchDIVER \cite{weart} as the local one. The Neural Network will be trained for interaction tasks with a table, and the results will be compared with the baseline Smith Predictor control strategy.
I sistemi di teleoperazione offrono un metodo sicuro per consentire agli esseri umani di interagire con ambienti precedentemente inaccessibili per diversi motivi di sicurezza. Si caratterizzano per un canale di comunicazione che collega il robot di comando a quello comandato. Un sistema di teleoperazione unilaterale è caratterizzato dall'invio di informazioni dal robot di comando a quello comandato, senza scambio di informazioni al contrario. In un sistema di teleoperazione bilaterale, questo scambio avviene da entrambe le parti, dove il robot comandato invia segnali (di solito informazioni sensoriali sull'ambiente remoto) all'operatore umano. Tuttavia, i canali di comunicazione possono introdurre ritardi temporali, specialmente in scenari in cui è presente una grande distanza tra le due parti, portando a instabilità nel sistema. In questo articolo, proponiamo una continuazione basata sui risultati di \cite{1626803}. Testiamo l'architettura proposta senza fare ipotesi sulla natura del ritardo temporale, su un sistema più complesso costituito da un robot Panda 7 DoF come robot remoto e un operatore umano equipaggiato con un WEART TouchDIVER come interfaccia locale. La rete neurale sarà addestrata per compiti di interazione con un tavolo, e i risultati saranno confrontati con la strategia di controllo basata sullo Smith Predictor.
Predictive force feedback for teleoperation systems under time delay
JETTI, GEORGES
2023/2024
Abstract
Teleoperation systems offer a safe method for humans to interact with environments that were previously unreachable due to several safety reasons. It is characterized by a communication channel connecting the commending robot to the commended one. A Unilateral Teleoperation system is characterized by information being sent from the commending robot to the commend one, only. In a bilateral Teleoperation system, this exchange happens from both sides, where the commended robot also sends signals (usually sensor information about the remote environment) to the human operator. However, communication channels can introduce time delays, especially in scenarios where a large distance is present between both parties, leading to instabilities in the system. In this paper, we offer a continuation based on the results of \cite{1626803}. We test the proposed architecture without making assumptions on the nature of the time delay, on a more complex system consisting of a Panda 7 DoF robot \cite{Franka} as the remote robot and a Human operator equipped with a WEART TouchDIVER \cite{weart} as the local one. The Neural Network will be trained for interaction tasks with a table, and the results will be compared with the baseline Smith Predictor control strategy.File | Dimensione | Formato | |
---|---|---|---|
2024_10_Jetti_Executive_summary.pdf
non accessibile
Descrizione: Executive summary
Dimensione
1.36 MB
Formato
Adobe PDF
|
1.36 MB | Adobe PDF | Visualizza/Apri |
2024_10_Jetti_thesis.pdf
non accessibile
Descrizione: Thesis
Dimensione
4.22 MB
Formato
Adobe PDF
|
4.22 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/10589/227086