b. Use Case with Claims

System: Social Robot (conversation partner, memory support)

Primary: Dementia Patient (socializer)

Secondary: Caregiver (human support)

• Robot detects a patient sitting alone or being inactive in a social setting (idle).
• Patient expresses boredom or lack of engagement (“I don’t know what to do”).
• Caregiver signals that a resident could benefit from social interaction.

• Environmental State: In the facility’s common lounge; several residents are present, engaged in light conversation or individual activities.
• Human State: Rebecca is seated alone and not participating in social interaction. The caregiver is assisting another resident and cannot intervene.
• System State: The robot is active and monitoring the environment. It has access to previously collected interest profiles from questionnaires and conversational data (all consented and locally stored). Facial and speech-recognition modules are active for detecting presence and affect.

MF1. Robot notices Rebecca alone and greets her (“Hi Rebecca, how are you today?”).

MF2. Robot asks light, open-ended questions to elicit interests  (“What kind of music do you like?”) or have small conversation ("How are you?")

MF3. Robot retrieves or updates Rebecca’s stored interests.

MF4. Robot cross-checks interests with nearby patients.

MF5. Robot facilitates an introduction (“Mary was just telling us about her honeymoon—you also love to travel!”).

MF6. Robot monitors conversation and provides light prompts if needed, then disengages once social interaction continues naturally.

AF1: Patient gives minimal answers → Robot probes gently with new topics or uses family-provided interest data.

AF2: No matching interests detected → Robot suggests group activity or transitions to general chat.

AF3: Patient expresses distress → Robot stops and alerts caregiver.

• Environmental State: The common lounge now has active conversation between Rebecca and another resident. Socializing has increased in general in the room.
• Human State: Rebecca is conversing comfortably, showing positive affect or laughter. The caregiver observes improved engagement without direct facilitation.
• System State: Robot logs the successful introduction, updates both residents’ interest profiles, and flags new social pairings for caregiver review. Robot returns to monitoring mode.

• If system/data is unavailable → Robot defaults to general conversation starter prompts.
• If patient consent is withdrawn → Robot disables access to personal data.
• If patient refuses engagement → Robot politely disengages and records the event.

System: Social Robot (activity scheduler and reminder)

Primary: Dementia Patient (activity participant)

Secondary: Caregiver (human support)

• The patient expresses boredom, inactivity is detected, or it is time for a scheduled activity.
• The caregiver updates or confirms available activities in the system.

• Environmental State: Several residents are seated in the common lounge of the care facility. No scheduled group activity is currently in progress.
• Human State: Rebecca, a PwD, appears idle and disengaged. Caregiver John is occupied with administrative tasks and not in the immediate area.
• System State: The robot is active and connected to the facility’s scheduling database. The system has detected an unstructured time period and has access to stored patient profiles, including Rebecca’s activity preferences and consent information.

MF1. Robot detects unstructured time or recognizes the patient’s idle state.

MF2. Robot retrieves data on available activities, locations, and required caregivers.

MF3. Robot cross-references patient interests and current availability.

MF4. Robot offers suitable activities using polite, supportive phrasing:

• “Would you like to get coffee with Maria in the lounge?”
• “There’s a group walk starting in 10 minutes—should I walk with you there?”
• “There’s a music group in the blue common area. Would you like to join?”

MF5. Patient responds verbally or through tablet interface.

MF6. Robot confirms choice, provides reminder when closer to activity, provides guidance or directions, and notifies caregivers if their presence is needed.

MF7. After the activity, robot asks for quick feedback: “Did you enjoy that?” or notes attendance/participation.

MF8. Robot updates patient preference data accordingly.

AF1: Patient declines → Robot offers gentle alternative (“Would you like to do something quieter, like reading or listening to music?”).

AF2: No available activities → Robot engages patient in conversation, suggests conversation with another patient, or suggests a solo task.

AF3: Patient is confused or unsure → Robot repeats the prompt more slowly or displays a visual option on its tablet. If needed, alerts a caregiver.

• Environmental State: An activity (e.g., coffee meeting or music session) is now in progress. The lounge area shows increased engagement and social interaction among residents.
• Human State: Rebecca has agreed to and joined an appropriate activity, expressing interest or enjoyment. Caregiver John has been automatically notified of participant status and required supervision, if applicable.
• System State: The robot has logged Rebecca’s participation, updated her preference data, and recorded feedback to refine future suggestions. The robot returns to standby or assists additional patients.

• If system/data is unavailable → Robot defaults to generic prompts (“Would you like to do something fun now?”).
• If consent is revoked → Robot ceases using personalized data and switches to neutral, non-specific suggestions.
• If patient is distressed or has a negative reaction -> Robot apologizes, ends interaction, and alerts caregiver.