Procurement teams evaluating eldercare service robots often begin with the vendor's demonstration screen. A clean interface, large buttons and readable contrast can look persuasive during a sales meeting, but those qualities belong to a specific domain: web content accessibility. W3C's Web Accessibility Initiative states that W3C accessibility standards address most older user needs for websites, web applications and web tools, and that accessible design for people with disabilities also helps older users with declining vision, physical ability, hearing and cognitive ability on the web (see Older Users and Web Accessibility). That scope is about web interaction, not about whether a mobile robot can safely approach a resident, recover from a stalled task or hand control back to a human caregiver. A procurement checklist should therefore separate screen accessibility from physical robot usability from the first evaluation stage.
W3C explains that WCAG materials include guidelines and techniques for making websites and web applications work better for people with disabilities and for older users with accessibility needs due to ageing (see Older Users and Web Accessibility). When a robot's onboard tablet or companion app meets WCAG success criteria, the vendor has shown that the digital layer follows web standards. The robot itself, however, operates in corridors, bedrooms and dining areas. During procurement observation, reviewers should treat the screen as one artefact and the machine's motion, sensing and recovery behaviour as a separate evidence layer. WCAG success criterion 2.5.3 Label in Name requires that for user interface components with visible text labels, the accessible name contains that visible text, a web-level rule for speech-input users (see WCAG 2.2 Label in Name). A hypothetical example may help: assume a robot's app has a button labelled "Call Nurse" whose screen-reader name says only "Emergency". A visually impaired resident using voice control might say "Call Nurse" and the interface would not match, even though the button looks correct. That hypothetical mismatch is a web-accessibility finding, not proof that the robot's navigation around a walker is unsafe. Procurement should record both layers separately.
Onboarding is the period when a resident first meets the robot. Reviewers should observe how instructions are delivered, whether the caregiver can pause the process and whether the robot asks for confirmation before moving. W3C notes that reduced short-term memory, difficulty concentrating and distraction can make it hard to follow navigation and complete online tasks, an overlap with disability needs on web interfaces (see Older Users and Web Accessibility). In a physical onboarding walkthrough, the equivalent observation is whether the robot repeats guidance, allows a slower pace and lets a staff member take over. A recommended documentation practice is to keep an onboarding log with the observed participant role, the time taken to complete first use and any moment when the caregiver interrupted the flow. This log is a procurement record rather than a claim about a health benefit.
Service robots meet dropped objects, closed doors and people who step into their path. Procurement observers should watch what the robot does when it cannot proceed: does it stop, explain the reason on screen, offer a simple next step and wait for a human? The web-side guidance from W3C covers notifications and feedback as a video topic for older users, concerning understandable content on the web (see Older Users and Web Accessibility). The physical counterpart is whether the robot's notification is legible in a dim corridor and whether the recovery action is gentle. Reviewers should record the error condition, the robot's visible response and whether a staff member resolved it.
Control handover is the moment when the robot yields authority to a caregiver or the resident. During procurement, observe whether the handover is announced clearly, whether the human can stop motion immediately and whether the robot confirms receipt of the new command. The visible-label and accessible-name relationship in WCAG 2.2, requiring the accessible name to contain the presented visible text, supports predictable interaction for speech and assistive technology on the interface (see WCAG 2.2 Label in Name). A recommended documentation practice is a handover checklist: who initiated the transfer, what signal appeared, how many seconds elapsed before the human had full control and whether the robot resumed only after explicit confirmation. This checklist is a procurement-observation tool, not a certification requirement of the cited W3C standard.
The W3C Web Accessibility Initiative develops web standards and focuses on making the web accessible to people with disabilities, with WCAG applying to websites and web applications rather than to mechanical certification (see Older Users and Web Accessibility). Procurement documents should therefore avoid wording that presents a WCAG-conformant screen as proof that the whole robot is certified for clinical use, safe in all homes or beneficial to health. An institutional source is not an endorsement of our checklist, and W3C does not evaluate robot hardware. Reviewers may write that the digital layer was observed against web accessibility guidance, while physical performance was observed in a controlled walkthrough.
Hypothetical example: suppose a care facility tests a delivery robot named Jazzy in a simulated lounge. The app meets the visible-label rule because the "Stop Jazzy" button's accessible name contains that visible text, satisfying the web criterion at the interface level (see WCAG 2.2 Label in Name). During observation, Jazzy encounters a chair moved into its path. It halts, shows "Path blocked, please clear or press Handover," and waits. A caregiver presses the handover control and Jazzy confirms "Control with caregiver." The procurement team records the screen-conformance result, the error condition, the notification wording and the handover time in one table. This hypothetical example shows how to keep web-screen evidence and physical-recovery evidence in the same file but in separate columns, without concluding that the robot is medically safe or legally compliant for every setting.
For further equipment reading, browse related industry updates and technical reference articles alongside this practical note.