TL;DR
Researchers in Singapore evaluated the HIRO (Healthcare-assistive Infection-control RObot), a multi-functional robot designed for sanitization, wayfinding, and health screening in a busy primary care clinic. The study found that over 90% of patients and healthcare workers viewed the robot’s functions favorably, leading to a design pivot from chemical disinfectants to UV-C light based on user safety concerns.
What Problem Does This Paper Address?
The COVID-19 pandemic highlighted a critical vulnerability in primary care: the need for rigorous, frequent infection control in high-traffic environments where manpower is already stretched thin. While robots offer a solution for "3D" jobs (Dull, Dirty, and Dangerous), their success depends entirely on Human-Robot Interaction (HRI).
If patients find a robot intimidating or if healthcare workers (HCWs) fear job replacement, the technology fails. This paper addresses the "acceptance gap" by investigating how a multidisciplinary group of users perceives a robot that monitors their temperature, checks their masks, and sanitizes their environment in a real-world clinical setting.
Methods
The researchers conducted a cross-sectional questionnaire survey at Tampines Polyclinic in Singapore.
- The Hardware: HIRO is a modular robot featuring an Enterprise Application Integration (EAI) mobile platform with SLAM navigation (LiDAR and IMU). It includes a thermal camera for fever screening, an RGB-D camera for mask detection, and a retractable "disinfection dome" equipped with UV-C lights.
- The Study Design: 385 participants (300 patients/visitors and 85 HCWs) were recruited. Because the robot was in a pilot phase, participants watched a 3-minute video of HIRO’s functionalities and were then given the option to interact with it directly.
- Data Collection: An e-survey gathered demographic data and feedback via 4-point Likert scales. The researchers used Chi-square tests to find associations between user demographics and their perceptions of the robot.
Datasets
The study did not use standard machine learning benchmarks. Instead, it generated a primary dataset of 385 survey responses from a diverse, multi-ethnic Asian population in Singapore.
- HCW Cohort: Included doctors, nurses, pharmacy staff, and allied health professionals.
- Public Cohort: Included patients and visitors of varying ages and educational backgrounds.
Key Insights and Ideas
- Feedback-Driven Iteration: Originally, HIRO was designed to spray liquid disinfectant. However, ~30% of participants expressed concerns about the health impact of inhaling these chemicals. This led the team to pivot to UV-C irradiation performed strictly after-clinic hours.
- The "Embarrassment Factor": While mask-detection was highly rated for utility, nearly 25% of HCWs and 14% of visitors found the robot’s voice-annotated instructions potentially "upsetting." This suggests that automated social policing can be perceived as intimidating or embarrassing in a public lobby.
- Gender and Perception: Women were statistically more likely than men to perceive that the robot improved the overall clinic experience.
- Trust vs. Interaction: Interestingly, while favorability was high (>90%), only 28.6% of participants chose to interact with the robot directly. This suggests that "passive acceptance" (being okay with the robot's presence) is easier to achieve than "active engagement."
Key Results
The quantitative favorability ratings for HIRO's core functions were remarkably high across both groups (Patients/HCWs):
- Sanitizing: 96.7% / 91.2%
- Mask Checking: 97% / 89.4%
- Temperature Monitoring: 97% / 91.7%
- Ushering/Wayfinding: 91.7% / 81.1%
- User Friendliness: 93% / 88.3%
- Overall Experience Improvement: 96% / 94.2%
Despite these high marks, a small but significant minority (10.6% of HCWs) had safety concerns, primarily regarding the risk of visitors damaging the robot or patients tripping over it.
Limitations and Future Work
- Language Bias: The survey was limited to English speakers, potentially excluding older segments of the population in a multi-lingual society like Singapore.
- Recruitment Bias: The "case-encounter" mode of recruitment might favor those already interested in technology.
- Job Displacement: Qualitative feedback revealed lingering concerns about robots replacing human staff. Future work needs to emphasize robots as "force multipliers" rather than replacements.
- Next Steps: The authors suggest future studies should include cost-benefit analyses, long-term exposure assessments (to see if the "novelty effect" wears off), and the implementation of AI-driven "staircase detection" to improve physical safety.
Summary updated