Usability and Acceptance of a Social Robot in People With Multiple Sclerosis: Comparative Cross-Sectional Study

Socially assistive robotics (SAR) is an emerging field integrating robotics, medicine, cognitive and social sciences, and neuroscience. Its potential in multiple sclerosis (MS) remains largely unexplored, although MS affects sensory-motor, cognitive, and psychological domains where SAR could offer substantial benefits.

This study aimed to evaluate the usability of a humanoid robot in people with MS, compare it with age-matched healthy individuals (HIs), and explore usability predictors. Automated written-language analysis was tested as a complementary approach.

In this 2025 cross-sectional study at the Italian MS Foundation laboratory, 20 people with MS (mean age 51.4, SD 11.0 years; n=13, 65% female; mean education 13.0, SD 3.4 years) and 20 age-matched HIs (mean age 47.3, SD 10.1 years; n=12, 60% female; mean education 14.8, SD 2.2 years) completed 4 predefined tasks with the robot Pepper. Usability was assessed using the System Usability Scale (SUS). Attitudes toward robots, robot-related factors, and task-related effort were measured with the General Attitudes Towards Robots Scale (GAToRS), Godspeed questionnaire, and NASA-TLX, respectively. Psychological variables (eg, personality traits from the Italian Ten-Item Personality Inventory [I-TIPI]) and clinical characteristics for people with MS (eg, cognitive functions via the app DIGICOG-MS, upper-limb difficulties with the Arm Function in Multiple Sclerosis [AMSQ]) were collected. Written responses to open-ended questions were analyzed using natural language processing (NLP).

SUS scores exceeded the standard cut-off in both groups, with no between-group differences (people with MS: mean 81.00, SD 12.31; HIs: mean 77.75, SD 11.24; r=0.14, 95% CI 0.01-0.42; P=.39), indicating high usability. People with MS reported more negative attitudes toward robots at the societal level (GAToRS S-: r=0.36, 95% CI 0.08-0.63; P=.02) and judged Pepper as more anthropomorphic (r=0.42, 95% CI 0.15-0.67; P=.007) and intelligent (r=0.37, 95% CI 0.09-0.61; P=.02) than HIs. Anthropomorphism was strongly associated with visuospatial memory (DIGICOG-MS: ρ=-0.632, 95% CI -0.84 to -0.26; P=.003) and relatively strong with both verbal memory (DIGICOG-MS: ρ=-0.504, 95% CI -0.77 to -0.08]; P=.02) and agreeableness (I-TIPI: ρ=0.491, 95% CI 0.06-0.77; P=.03). Exploratory forward stepwise regression showed that conscientiousness (I-TIPI: B=2.064, 95% CI 0.679-3.449), negative personal attitudes toward robots (GAToRS P-: B=0.169, 95% CI 0.033-0.306), and perceived upper-limb difficulties (AMSQ: B=0.169, 95% CI 0.033-0.306) explained 64% of SUS variance (adjusted R²=0.568); task-related frustration (NASA-TLX: B=-1.497, 95% CI -2.276 to -0.719) explained 48% (adjusted R²=0.451). NLP analyses revealed no between-group differences in lexical productivity or sentiment-lexicon coverage.

This study provides first preliminary evidence that people with MS perceive a social robot as highly usable, comparably to HIs, suggesting MS-related challenges may not hinder human-robot interaction. Psychological factors, prior attitudes toward robots, and clinical characteristics were associated with user experience in the MS group. Future research should explore whether integrating these measures with language-based approaches informs tailored interventions in MS.