AAU Student Projects - visit Aalborg University's student projects portal
A master's thesis from Aalborg University
Book cover


Exploring User Interactions with the Articulated Head in Different Settings

Authors

;

Term

4. term

Publication year

2019

Submitted on

Pages

101

Abstract

This thesis investigates how people interact with the social robot Articulated Head (AH) across different settings, aiming to design convincing and engaging museum interactions and to assess how far laboratory findings transfer to real-world environments. The work comprises two grounding studies (interviews and walk-alongs with visitors at Questacon’s Born or Built? exhibition, and a lab study examining how two base positions affect users’ perception of the robot) and two main studies with an identical research design conducted in a controlled laboratory and in the wild at a university event. Using a mixed-methods approach, we collected video and observational data, conducted pre- and post-interaction interviews, and administered Godspeed questionnaires; quantitative data were interpreted for trends due to a small sample, and qualitative data were analyzed thematically. Across studies, we derived six concrete design suggestions to strengthen interaction with AH in a museum context: place the robot’s base in a profile position; use eye contact to indicate what the robot is attending to; position the interaction in front of the robot so people stand face-to-face; remove the animated neck and shoulders from the screen; anticipate that visitors will touch the robot and plan safety measures; and use the virtual head’s animations to signal forthcoming movements to support predictability. Comparing the lab and field studies revealed many similarities alongside context-specific differences; participants’ expectations appeared higher in the lab and may have contributed to lower ratings. Overall, the findings indicate that some HRI results can transfer from laboratory to real-world settings to some extent, though larger samples and further studies are needed. The work provides practical guidance for advancing AH and methodological direction for understanding, predicting, and designing HRI across settings.

Dette speciale undersøger, hvordan mennesker interagerer med den sociale robot Articulated Head (AH) i forskellige omgivelser, med fokus på at designe en overbevisende og engagerende museumsoplevelse og vurdere, i hvilket omfang laboratoriefund kan overføres til virkelige miljøer. Projektet omfatter to forberedende studier (interviews og walk-alongs blandt besøgende på Questacons Born or Built?-udstilling samt et laboratoriestudie af, hvordan to basepositioner påvirker perceptionen af robotten) og to hovedstudier med samme forskningsdesign gennemført i hhv. et kontrolleret laboratorium og i felten ved et universitetsarrangement. Med en mixed-method tilgang indsamlede vi video- og observationsdata, gennemførte interviews før og efter interaktionen og anvendte Godspeed-spørgeskemaer; de kvantitative data blev analyseret for tendenser pga. et lille datasæt, mens de kvalitative data blev temaanalyseret. På tværs af studierne fremkom seks konkrete designforslag til at styrke interaktionen med AH i museumsregi: placer robotbasen i profil; brug øjenkontakt som signal om, hvad robotten retter opmærksomheden mod; placer interaktionen foran robotten, så bruger og robot står ansigt til ansigt; fjern den animerede hals og skuldre på skærmen; forudse at mennesker vil røre robotten og planlæg sikkerhedstiltag; og brug det virtuelle hoveds animationer til at signalere næste bevægelse, så brugerne bedre kan forudsige robotten. Sammenligninger mellem laboratorie- og feltstudiet viste mange sammenfald, men også kontekstspecifikke forskelle; deltagernes forventninger syntes højere i laboratoriet og kunne bidrage til lavere bedømmelser. Overordnet peger resultaterne på, at visse HRI-fund kan overføres fra laboratorium til virkelige omgivelser i et vist omfang, men der er behov for større stikprøver og flere studier. Arbejdet giver praktiske anbefalinger til videreudvikling af AH og metodiske anvisninger til at forstå, forudsige og designe HRI på tværs af settings.

[This apstract has been generated with the help of AI directly from the project full text]