Care robotics is steadily moving from controlled lab environments into everyday care settings such as rehabilitation spaces and assisted living facilities. But introducing robots into care environments raises an important question:
How do we scale robotic support in care without compromising dignity, safety and quality of care?
Today, many robotic systems rely primarily on cameras to perceive their surroundings. While vision is an important source of information, robot cameras alone cannot and should not capture everything. Physical strain, fatigue or increasing effort are often invisible, yet these can be important indicators during exercise or rehabilitation. At the same time, continuous camera monitoring can feel intrusive, particularly in places where people expect privacy.
Wearables offer a different perspective. They are increasingly used in everyday life and can provide additional signals about movement and physiological response that cameras cannot see. However, asking people to continuously share raw biometric data with robotic systems creates another challenge. Even if the intention is to improve care, many people are understandably uncomfortable with sharing such personal information continuously.
Created with AI by Google Gemini
This creates a tension between what robotic systems would ideally sense and what users consider acceptable in care environments. To better understand how these aspects can be brought into balance in real-world settings, we conducted stakeholder interviews within the Bio-Link project as part of the FORTIS Project1 initiative. The aim was to explore how older adults and care professionals perceive assistive technologies in everyday care, and what they consider essential for acceptable and trustworthy use.
Privacy
Privacy was a central concern throughout the discussions. Participants emphasised the need for clarity about what data is collected, where it is processed, who can access it, and when systems are active. This reinforced our privacy-by-architecture approach, where data minimisation, explicit consent, and local processing are not add-ons but core system principles.
Easy to use technology for wellbeing
The second important finding was that participants were generally open to assistive technologies, as long as they integrate naturally into daily routines and do not introduce additional complexity. Systems need to feel intuitive and clearly useful in the moment of interaction. At the same time, participants clearly distinguished between technology for everyday wellbeing and technology associated with illness or frailty. While everyday applications, such as activity tracking or vital sign monitoring, were widely accepted, systems linked to frailty or loss of independence were viewed more critically.
Control & Transparency
A further cross-cutting insight was that trust is not only determined by technical performance, but by perceived control and predictability. Participants expressed a preference for systems that make their sensing and decision-making transparent, and that allow users or caregivers to easily adjust the system to their preferences. Adaptive and context-aware sensing was seen as more acceptable, especially when users could understand when and why a system becomes active. This points towards hybrid perception strategies that combine minimal necessary sensing with explicit interaction cues and user-governed transparency.
Created with AI by Google Gemini
In summary, the findings highlight that the successful integration of robotic systems in care settings depends less on maximising sensing capability and more on carefully balancing information needs with social and ethical acceptability. Designing for care contexts therefore requires not only technical robustness, but also a strong emphasis on transparency, adaptability and respect for personal boundaries in everyday environments.
1 The FORTIS Project is a European Union funded research initiative under the Horizon Europe programme focused on advancing human-robot interaction (HRI). It aims to develop technologies that enable robots and people to collaborate naturally, safely, and effectively over extended periods in real-world workplaces.
