TEDIRO attending the 35th rehab colloquium

It was an inspiring few days filled with many engaging discussions at the German Pension Insurance Rehabilitation Science Colloquium, held March 24–26, 2026, in Leipzig/Germany.

Our highlights
🔹 Live demonstrations of our mobile therapy assistance robot – accompanied by many engaging expert discussions.
🔹 Our two presentations outlined below in the session "Robotics, AI, and Prosthetics in Musculoskeletal Rehabilitation" during which most of the hall’s 250 seats were occupied.

Acceptance study
Christian Sternitzke presented the results of a clinical study on the use of a mobile robot in self-training. He outlined the conditions that determine the acceptance of AI-based systems, including the effects of algorithm aversion, and compared patients’ assessments of self-training alone, with a robot, and with therapists. His take-home message: Patients accept robot-assisted gait training without the presence of therapeutic staff as part of physical therapy. However, one audience member also succinctly summed it up: it is no longer a question of whether to use a robot or a therapist, but rather whether therapy can be offered at all.

AI-based model development
Anke Mayfarth provided insight into the scientific approach to AI-based analysis of movement sequences and explained how the ground truth for the training data was collected. It became clear how crucial physiotherapeutic expertise is for the development of reliable algorithms: TEDIRO systematically incorporates therapeutic expertise into its evaluation models, thereby laying the foundation for medically meaningful, AI-supported movement analysis.
This was particularly evident in the error classification during gait training with forearm crutches: The five therapists who labeled the training data for the ground truth agreed completely in only 2% of all cases.
This result in particular underscores just how challenging the assessment of movement quality is—and how important it is to digitize therapeutic expertise in a structured way.
The key insight here: The evaluation module on THERY is the digitized form of physiotherapeutic expertise—a colleague who never tires of providing valuable feedback during self-training and enables additional activation outside of limited therapy sessions.