Using Synthetic Data for Wheelchair Footrests Design Customization: The Kyklos 4.0 Approach


  • Patrick Lowin Fraunhofer Fokus, Germany
  • Yacine Rebahi Fraunhofer Fokus, Germany
  • Benjamin Hilliger Fraunhofer Fokus, Germany
  • Bowen Zheng Fraunhofer Fokus, Germany



deep learning, design customization, weelchair, footrest, synthetic datasets, Kyklos 4.0, circular economy


Wheelchairs are complex systems often requiring a wide range of adjustments to adapt to the various types of patients’ disabilities. They also include a series of additional elements, for example footrests, designed to keep the patient in a comfortable position. Unfortunately, most of the commercial products do not allow maintaining the position of a patient foot who has no control over his lower limbs. To address this issue, customization seems to be the appropriate solution as it enables to tailor products based on predetermined features. In (Gharra et al, 2023), we have explored the use of computer vision and AI to correctly define customized parameters of the wheelchairs’ footrests.  The proposed solution is based on estimating geometric properties of real shoes contours. Although this solution was accurate to some extent, its main drawback was the small amount of data that we were able to collect. For this reason, we decided to explore another approach where shoes contours data is synthetic, and Conventional Neural Networks (CNNs) are applied. This paper discusses the synthetic data approach and compares its performance to the one described in (Gharra et al, 2023).


Received: 30 August 2023 | Revised: 22 September 2023 | Accepted: 21 November 2023


Conflicts of Interest

The authors declare that they have no conflicts of interest to this work.


Metrics Loading ...




How to Cite

Lowin, P., Rebahi, Y., Hilliger, B., & Zheng, B. (2023). Using Synthetic Data for Wheelchair Footrests Design Customization: The Kyklos 4.0 Approach. Artificial Intelligence and Applications.



Online First Articles