Development and Control of a Robotic Assistant Walking Aid for Fall Risk Reduction

Marcel Naderer¹Yeongjae Kim²Tae-Hyoung Kim^2*Yeongmi Kim^1*

• ¹MCI Management Center Innsbruck, Innsbruck, Austria
• ²Chung-Ang University, Dongjak-gu, Republic of Korea

Falls are a major risk factor among the elderly, often resulting in injuries that compromise independence and quality of life. Conventional walking aids lack active stabilization capabilities and are therefore limited in effectively preventing balance-related accidents. This paper presents the design and control of a smart robotic assistant aimed at reducing fall risk in elderly users by providing real-time balance support. The proposed system uses a wearable inertial measurement unit to detect postural imbalances in the sagittal (front-back) and frontal (side-to-side) planes.When instability is detected, the robotic arm generates compensatory forces or torques through linear or rotational actuators to help the user regain a stable posture. Using a cascaded control architecture, the outer loop is designed to maintain the user's upright posture, while the inner loop ensures fast and accurate actuator performance. To enable effective and reliable control in the real system, actuator dynamics are characterized through an optimization-based system identification approach, resulting in transfer function models with over 98 % accuracy. Based on these models, PID controllers are optimally tuned using an optimization algorithm to ensure fast and accurate corrective action. The system effectively returns the user to a stable position within 2.3 ± 0.3 s for linear actuation (with a response time of 120 ± 10 ms) and 2.2 ± 0.2s for rotary actuation (with a response time of 140 ± 15 ms), providing safe posture return during imbalance events. To further enhance safety, an automatic braking mechanism immobilizes the walking aid during corrective maneuvers. Experimental validation demonstrates the system's effectiveness in detecting and correcting postural imbalances in both the sagittal and frontal planes under dynamic conditions. These results highlight the potential for enhancing mobility, safety, and therapeutic support for older adults, contributing to the advancement of assistive fall-prevention technologies.