Six Questions to Ask a Researcher About Care Robots in Nursing Homes

What care robots actually do

Care robots are physical machines designed to assist with activities associated with health, everyday living or social interaction. In nursing homes, the category includes socially assistive robots that encourage conversation or participation, interactive robotic animals, mobile devices that support selected tasks and humanoid systems that provide information or structured activities. Some use artificial intelligence, while others operate through relatively simple programmed responses. Calling all of them AI companions obscures important differences in their capabilities.

Social robots for elderly companionship are one particularly visible part of this landscape. A robotic animal may respond to touch, movement or sound, creating an activity that residents can engage with individually or in a group. A conversational robot may offer games, prompts or scheduled information. Neither design automatically provides the understanding, judgment or responsibility associated with a human caregiver.

A useful distinction is between a robot that provides an activity and one that performs a care task. Encouraging a resident to participate in a conversation is different from administering medication, assessing a medical condition or making a decision about personal care. Those distinctions should be clear in the system's intended purpose and the facility's operating procedures.

• Social interaction: conversation, prompts and shared activities.
• Emotional engagement: responsive robotic animals and familiar interaction.
• Cognitive activities: games, memory prompts and structured exercises.
• Everyday assistance: selected reminders, information and practical support.
• Physical support: specific rehabilitation or movement-related activities.

Who is the technology intended to serve?

The resident is often the most visible user, but a nursing home robot operates within a wider care environment. Residents may have different communication preferences, mobility needs, sensory abilities, cultural backgrounds and levels of familiarity with technology. One person may enjoy a responsive robotic animal, while another may find its movements or sounds uncomfortable. A third may prefer a human-led group activity. Personal preference is an important part of determining whether an interaction is appropriate.

Care staff are users too. They may need to introduce the device, explain its purpose, prepare activities, supervise interactions, clean equipment and resolve technical problems. Family members may also have questions about why a robot is being used and what information it collects. A system that appears simple during a demonstration can create additional work when these responsibilities are not planned.

Research on robots in assisted living has identified a recurring difficulty: individual responses do not tell the whole story. A 2023 scoping review examined 73 publications representing 69 studies. Only 18 of those studies considered the broader care context. This matters because a positive response from a resident does not by itself establish that a technology fits the needs of an entire facility.

The six questions that clarify the category

A useful discussion with a researcher begins with six questions. They are also a practical framework for understanding a particular device, its intended role and the evidence surrounding it.

1. What does it do? Ask for the actual activities the robot supports. Does it respond to touch, facilitate group conversation, provide reminders or lead a structured exercise? Request a distinction between demonstrated functions and proposed future capabilities.

2. Who is it for? Ask which residents the design considers, whether participation is voluntary and how accessibility, language, sensory preferences and individual routines are addressed.

3. What does it not do? Ask about tasks outside its intended role. A social robot is not automatically a therapist, nurse, emergency responder or independent decision-maker.

4. What happens when it ends? Ask what happens during a power failure, maintenance period, software change or permanent withdrawal. Consider whether residents lose access to a familiar activity and whether alternatives are available.

5. Who decides what it says? Ask whether conversations follow fixed scripts, staff-selected content, remote configuration or AI-generated responses. Clarify who is responsible for correcting inappropriate information.

6. What should a user know first? Ask about privacy, consent, supervision, physical accessibility, cleaning, costs and the amount of staff time required.

What the research can and cannot establish

Research into care robotics covers several different outcomes. Researchers may examine enjoyment, willingness to participate, interaction frequency, perceived companionship, usability or changes in a particular activity. These are not interchangeable measures. A resident enjoying a session is meaningful information about the experience, but it does not establish a lasting improvement in health or a reduction in loneliness.

A 2024 scoping review of AI-enabled robots in long-term care homes identified three broad implementation barriers: technical complexity and limitations, concerns about usefulness and ethical consequences, and insufficient resources. These findings highlight the importance of the setting. Internet access, available space, staff training, maintenance arrangements and equipment costs all influence whether a robot can become part of everyday care.

The evidence also needs to be interpreted according to the type of robot and the length of observation. Results involving a robotic animal cannot automatically be applied to a humanoid conversational system. Short-term engagement does not answer questions about long-term use, and findings from one facility may not transfer to another with different staffing, residents or routines.

What happens around the robot

The central design question is not simply whether a robot can interact with a resident. It is how that interaction fits into the resident's day. A device might encourage a shared activity, provide a familiar object for conversation or make an individual session more engaging. Its contribution depends on the circumstances in which people use it and whether the activity supports an existing care goal.

Consider a hypothetical nursing home introducing a responsive robotic animal for optional group sessions. A meaningful evaluation would consider who chooses to participate, whether residents enjoy the activity, how staff supervise it, what happens when someone becomes uncomfortable and whether the equipment is available when expected. It would also consider whether the activity complements other opportunities for social interaction.

This approach keeps responsibility visible. A robot may facilitate an encounter, but staff and care organizations remain responsible for the surrounding environment and the decisions that affect residents. Questions about consent, privacy and individual care needs should be addressed through the facility's established professional and organizational procedures.

A practical starting point for families and care teams

Anyone encountering a care robot in a nursing home can begin by asking the facility to describe its intended purpose in one sentence. Then ask who introduces it, whether residents can decline, what information it records and who maintains it. These questions reveal more about the actual role of a device than a general claim that it provides companionship.

The next step is to request the facility's explanation of supervision and alternatives. If a resident does not enjoy the interaction, there should be another activity available. If the robot stops working, the care plan should not depend on its continued operation. Families with questions about clinical suitability should speak with the appropriate care professional, while technical and privacy questions should go to the organization responsible for the device.

Questions readers ask