How can AI-based movement monitoring support rehabilitation after hip fracture without compromising older adults’ autonomy and person-centred care?

I am developing a doctoral research direction involving AI-assisted movement monitoring for older adults recovering after hip fracture. The intended role is to identify movement patterns or fall-risk indicators and support timely, individualized rehabilitation—not to replace clinical judgment or human interaction.

What design features and outcome measures would best demonstrate that such a system is clinically useful, acceptable to older adults, and ethically implemented? I am particularly interested in perspectives on usability, privacy, consent, equity, clinician oversight, functional recovery, fear of falling, and adherence to rehabilitation.

Insights from rehabilitation clinicians, geriatric-care researchers, AI developers, and health-ethics experts would be very welcome.