Benjamin Filtjens Human Movement Intelligence

Research project

AI@WZC

A co-designed robotic exercise assistant for care-home residents where staff time for exercise supervision is limited.

Research Collaboration Rehabilitation and assistive technologyMovement in clinical practice
Role
Co-Investigator
Award / reference year
2023
Funders
Province of Flemish Brabant, Smart Region
Three illustrated co-design sessions: older care-home residents and researchers discuss exercise needs, explore ideas alongside a robotic assistant, and try the robot in a seated exercise session.

Problem understanding, ideation and pilot testing. Illustrative reconstruction with fictional participants and simplified settings.

Illustration developed for AI@WZC · View full-size image ↗

Co-designing a robotic exercise assistant

Care-home residents often spend much of the day sedentary, which can harm health and reduce independence. Exercise therapy can help maintain function, but limited care staff makes individual supervision difficult to provide.

AI@WZC explored a robotic assistant to support exercise in this setting. Residents and researchers co-designed the assistant across three sessions:

  1. Problem understanding: clarifying residents' needs and challenges around exercise.
  2. Ideation: exploring ideas for the assistant together.
  3. Pilot testing: trying the robotic assistant with residents in the care-home setting.

Related research developed a single-camera method to recognise exercises, estimate joint movements and derive measures such as repetition count, exercise duration and range of motion. The PLOS Digital Health study evaluates this approach for exercise assessment in older care-home residents.

Related work

Publications

2026

Journal Article

Deep learning-based physical exercise assessment of older adults using single-camera videos

Stefanova V, Maeyens E, Brughmans J, Vargemidis D, Lebleu J, Stulens L, Broeder S, Gilis K, Deschodt M, Vanwanseele B, Beckwée D, Gilat M, Filtjens B, Vanrumste B

Joint senior authors: Filtjens B and Vanrumste B. Corresponding author: Filtjens B.

PLOS Digital Health 5 (8), e0001576, 2026

Citation

Stefanova V, Maeyens E, Brughmans J, Vargemidis D, Lebleu J, Stulens L, Broeder S, Gilis K, Deschodt M, Vanwanseele B, Beckwée D, Gilat M, Filtjens B, Vanrumste B (2026). Deep learning-based physical exercise assessment of older adults using single-camera videos. PLOS Digital Health 5 (8), e0001576, 2026. https://doi.org/10.1371/journal.pdig.0001576 Joint senior authors: Filtjens B and Vanrumste B. Corresponding author: Filtjens B.