Team-Project: Development of a Standardized Testbench for Patient Bed Fall Detection and Prevention Systems

Institut

TUM Munich Institute of Robotics and Machine Intelligence (Institut)

Typ

Semesterarbeit /

Inhalt

experimentell /  

Beschreibung

Development of a Standardized Testbench for Patient Bed Fall Detection and Prevention Systems

Background
Patient falls are among the most severe postoperative incidents in hospitals, nursing or retirement homes, often resulting in serious injuries such as head trauma, fractures, or prolonged recovery times. While new AI-based systems are being developed to predict and detect fall events, their validation remains challenging, as realistic testing with patients is neither practical nor ethically acceptable.
In this project, TUM University Hospital and TUM MIRMI will jointly develop a standardized and repeatable testbench for evaluating patient fall detection and prevention systems. The setup will combine a hospital bed, a crash-test dummy with interchangeable neck designs, and a robotic system capable of generating realistic and reproducible fall scenarios under controlled laboratory conditions.
Project Structure
The project is designed as a 3-month full-time team internship and is ideally suited for students from programs such as Biomechanics, Robotics and Mechatronics. Three students will collaborate while each taking responsibility for a dedicated work package:
·      Robotics and Test System Development: Integration of the robot, patient bed, and crash-test dummy, including development of repeatable fall scenarios and experimental procedures.
·      Medical and Biomechanical Assessment: Investigation of clinically relevant fall scenarios, associated injury mechanisms, and injury risk metrics.
·      Sensor Benchmarking and Data Analysis: Evaluation of fall detection technologies, including a sensorized mattress and a radar-based monitoring system, against motion-tracking ground truth data.
The project will take place in the department of Orthopaedics at TUM University hospital and supervised by Christina Valle and Robin Kirschner to address the medical, AI-detection based, and Biomechanical/Robotics perspectives.
Project Tasks
The project includes:
·      Literature review on patient falls, injury mechanisms, and existing detection technologies.
·      Design of a standardized experimental methodology and test protocol.
·      Integration and calibration of the fall detection sensing
·      Setup and integration of the robotic test environment.
·      Execution of a pilot study involving approximately 30 fall experiments to assess repeatability and system performance.
·      Benchmarking of different sensing technologies with respect to detection accuracy and latency.
·      Comparison of injury estimates obtained using a conventional dummy neck and a newer biofidelic neck design.
Expected Outcomes
The project will deliver a functional robotic testbench for patient bed-fall research, quantitative performance data for different fall detection systems, and insights into the influence of crash-test dummy biofidelity on injury assessment. The results will contribute to the development of future benchmarking and validation methodologies for patient safety technologies in healthcare.
This project offers a unique opportunity to work as a team at the intersection of robotics, biomechanics, AI-based sensing, and clinical safety, while contributing to technologies that can help reduce fall-related injuries in hospitals.

Please apply as a team of 3 that want to cover the different aspects.

Voraussetzungen

- self-sufficient programming work (ideally knowledge of ROS)

- ideally experience in experimental procedures

- basic statistics knowledge

- good team-work skills

Möglicher Beginn

sofort

Kontakt

Robin Jeanne Kirschner
robin-jeanne.kirschnertum.de