This thesis investigates the interplay between Human Factors and the application of collaborative robots (cobots). Specifically, by employing the Tower of Hanoi as a tool, the experiments conducted aimed to explore how the cobot's size, speed, and modes of collaboration affect trust, anxiety, mental workload, understanding of the task, and performance. The experiments for the analysis were carried out at SUPSI in Lugano and involved 29 volunteers who had to solve the mathematical game of the Tower of Hanoi together with a collaborative robot (cobot). The cobot models used are ABB GoFa 15000 and igus ReBeL. For the execution of each experiment, the work developed a system integrating a vision system and detection of game elements, a graphical user interface (GUI), and an orchestrator coordinating the cobot operations. Additionally, a detailed execution protocol was developed, along with a questionnaire inspired by existing literature and a range of analysis techniques. This not only facilitated the experiments outlined in this study but also paves the way for future research and diverse analytical approaches. The experiments compare the different dimensions of the cobots, different linear speeds, and distinct collaboration modes, thus evaluating the impact of each on the volunteers' Human Factors. For each participant, data on their levels of trust, anxiety, mental workload are collected with subjective assessment tests (PTM Scale, NASA-TLX, TPS-HRI, STAI) as well as through wearable devices, and then analysed using parametric and non-parametric statistical tests (ANOVA, t-test, MWU test, KWH test). The results of these tests show that in the presence of a human-robot collaboration environment with a high perception of safety, the size and speed of the cobot do not significantly impact trust, anxiety and mental workload perceived while cooperating with the robot, concerning the collaboration mode. However, with varying modes of collaboration between humans and robots, significant differences are noted for collaboration with voluntary operator input: in such cases, mental workload is lower, along with a greater sense of "well-being," albeit at the expense of longer task completion times. Furthermore, for every alternated collaboration mode, trust and propensity to trust greatly increase, underlying how in situation of perceived safety the way that human and robot interact has a greater effect with respect to physical characteristics of the robot such as size and speed.
Questo lavoro di tesi indaga l'interazione tra i Fattori Umani e l'applicazione dei robot collaborativi (cobots). In particolare, attraverso l'utilizzo della Torre di Hanoi, gli esperimenti condotti mirano a esplorare come le dimensioni, la velocità e le modalità di collaborazione del cobot influenzino la fiducia, l'ansia, il carico mentale, la comprensione del compito e le prestazioni. Gli esperimenti per l'analisi sono stati condotti presso la SUPSI a Lugano e hanno coinvolto 29 volontari che hanno dovuto risolvere il gioco matematico della Torre di Hanoi insieme a un robot collaborativo (cobot). I modelli di cobot utilizzati sono ABB GoFa 15000 e igus ReBeL. Per l'esecuzione di ciascun esperimento, è stato sviluppato un sistema che integra un sistema di visione e rilevamento degli elementi di gioco, un'interfaccia utente grafica (GUI) e un orchestratore che coordina le operazioni del cobot. Inoltre, è stato sviluppato un protocollo di esecuzione dettagliato, insieme a un questionario ispirato alla letteratura esistente e a una serie di tecniche di analisi. Ciò non solo ha facilitato gli esperimenti descritti in questo studio, ma ha anche aperto la strada a futuri studi e a diverse approcci analitici. Gli esperimenti confrontano le diverse dimensioni dei cobot, diverse velocità lineari e diverse modalità di collaborazione, valutando così l'impatto di ciascuna su i Fattori Umani dei volontari. Per ciascun partecipante, sono stati raccolti dati sui loro livelli di fiducia, ansia, carico mentale con test e questionari dalla letteratura oltre che attraverso dispositivi indossabili, e quindi analizzati utilizzando test statistici parametrici e non parametrici (ANOVA, test t, test MWU, test KWH). I risultati di questi test mostrano che in presenza di un ambiente di collaborazione uomo-robot con una percezione elevata di sicurezza, le dimensioni e la velocità del cobot non influenzano significativamente la fiducia, l'ansia e il carico mentale percepiti durante la cooperazione con il robot, riguardo alla modalità di collaborazione. Tuttavia, con diverse modalità di collaborazione tra esseri umani e robot, si notano differenze significative per la collaborazione alternata uomo-robo con l'input volontario dell'operatore: in tali casi, il carico mentale è inferiore, insieme a un maggior senso di "benessere", sebbene a discapito di tempi di completamento del compito più lunghi. Inoltre, per ogni modalità di collaborazione alternata, la fiducia e la propensione a fidarsi aumentano notevolmente, evidenziando come in situazioni di percezione di sicurezza il modo in cui umani e robot interagiscono abbia un maggiore effetto rispetto alle caratteristiche fisiche del robot come dimensioni e velocità.
Human-robot interaction: effects on human factors during a Tower of Hanoi collaboration task
PLONER, GIOVANNI
2022/2023
Abstract
This thesis investigates the interplay between Human Factors and the application of collaborative robots (cobots). Specifically, by employing the Tower of Hanoi as a tool, the experiments conducted aimed to explore how the cobot's size, speed, and modes of collaboration affect trust, anxiety, mental workload, understanding of the task, and performance. The experiments for the analysis were carried out at SUPSI in Lugano and involved 29 volunteers who had to solve the mathematical game of the Tower of Hanoi together with a collaborative robot (cobot). The cobot models used are ABB GoFa 15000 and igus ReBeL. For the execution of each experiment, the work developed a system integrating a vision system and detection of game elements, a graphical user interface (GUI), and an orchestrator coordinating the cobot operations. Additionally, a detailed execution protocol was developed, along with a questionnaire inspired by existing literature and a range of analysis techniques. This not only facilitated the experiments outlined in this study but also paves the way for future research and diverse analytical approaches. The experiments compare the different dimensions of the cobots, different linear speeds, and distinct collaboration modes, thus evaluating the impact of each on the volunteers' Human Factors. For each participant, data on their levels of trust, anxiety, mental workload are collected with subjective assessment tests (PTM Scale, NASA-TLX, TPS-HRI, STAI) as well as through wearable devices, and then analysed using parametric and non-parametric statistical tests (ANOVA, t-test, MWU test, KWH test). The results of these tests show that in the presence of a human-robot collaboration environment with a high perception of safety, the size and speed of the cobot do not significantly impact trust, anxiety and mental workload perceived while cooperating with the robot, concerning the collaboration mode. However, with varying modes of collaboration between humans and robots, significant differences are noted for collaboration with voluntary operator input: in such cases, mental workload is lower, along with a greater sense of "well-being," albeit at the expense of longer task completion times. Furthermore, for every alternated collaboration mode, trust and propensity to trust greatly increase, underlying how in situation of perceived safety the way that human and robot interact has a greater effect with respect to physical characteristics of the robot such as size and speed.File | Dimensione | Formato | |
---|---|---|---|
2024_04_Ploner_executiveSummary.pdf
Open Access dal 19/03/2025
Descrizione: executive summary della tesi
Dimensione
3.45 MB
Formato
Adobe PDF
|
3.45 MB | Adobe PDF | Visualizza/Apri |
2024_04_Ploner_Tesi.pdf
Open Access dal 19/03/2025
Descrizione: testo della tesi
Dimensione
14.96 MB
Formato
Adobe PDF
|
14.96 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/10589/218286