Background
As population aging and the shortage of caregiving manpower become increasingly serious, the demand for mobility assistance, rehabilitation training, and patient transfer support in medical and long-term care settings is rapidly increasing. Elderly individuals or patients with limited mobility often require physical support from nurses or caregivers during walking, standing up, obstacle crossing, and posture transitions. Over time, these tasks can place a significant burden on caregivers’ lower backs and limbs, and may even lead to occupational injuries.
Although many existing humanoid robots are already capable of walking, obstacle avoidance, and object manipulation, their development has mainly focused on imitating human motion, maintaining self-balance, and performing general-purpose tasks. They still show clear limitations in meeting medical care needs such as supporting the human body, assisting patients with walking, and ensuring safe physical interaction. In particular, when interacting physically with the human body, a robot must not only be able to stand and move stably, but also possess sufficient load-bearing capability, compliant control ability, and real-time responsiveness to the user’s movement intentions.
Therefore, this study targets medical-assistive humanoid robots as the application scenario and focuses on the development of bionic mechanisms, gait control, and human–robot collaborative interaction technologies for the lower-limb robotic leg system. The goal is to enable the robot to assist patients in walking, support, obstacle crossing, and posture transition tasks in rehabilitation and care settings, thereby reducing the caregiving burden and improving patient mobility safety.
Research Objectives
The main objective of this study is to develop a bionic robotic leg system suitable for medical assistance and rehabilitation care settings, enabling stable load-bearing, dynamic walking, human–robot coordination, and safe assistance. The specific objectives are as follows:
First, this study aims to establish a bionic leg mechanism with high load-bearing capability. Through crank-linkage and bionic mechanism design, the burden on motors and reducers can be reduced, while output torque and support capability can be enhanced, making the robotic leg more suitable for walking assistance and patient support tasks.
Second, this study aims to develop compliant posture control technology. With a variable link-length design, the robotic leg can adjust its posture according to terrain, load, and user conditions, supporting level-ground walking, obstacle crossing, and movement over rough terrain.
Third, this study will integrate force sensing and feedback control technologies to establish a safe human–robot interaction control framework. Through joint force/torque sensing, ground reaction force estimation, and impedance control, the robotic leg can provide stable and compliant support during interaction with patients.
Fourth, this study aims to complete walking-assistance and rehabilitation support tasks. Ultimately, the robotic leg will be integrated with upper-body mechanisms, arm support, and sensing systems to demonstrate application scenarios such as assisted walking.
Methods
This study will follow a technical approach centered on bionic mechanism design, gait planning, dynamic modeling, force feedback control, and experimental validation.
In terms of mechanism design, the robotic leg will adopt a bionic linkage and crank-driven architecture to emulate the support and propulsion characteristics of biological lower limbs during walking. Through linkage geometry design and force transmission analysis, motor output can be effectively converted into supporting and propulsive forces, reducing the burden on the actuation system and improving motion efficiency under high-load conditions.
In terms of gait planning, gait generation methods suitable for walking-assistance tasks will be developed. Factors such as stride length, step height, foot-end trajectory, center-of-mass transfer, and ground contact conditions will be considered, enabling the robotic leg to generate stable and adjustable walking patterns under different usage scenarios. To address the needs of rough terrain, a variable link-length mechanism will be further introduced to enhance the robot’s environmental adaptability.
In terms of control methods, motion control, force feedback control, and impedance control will be integrated so that the robotic leg can achieve both stability and compliance during load-bearing and interaction. Based on joint position, velocity, torque, and contact force information, the system will adjust the lower-limb output force and gait state to prevent excessive contact force or unstable support from causing discomfort or risk to the user.
In terms of system integration, the robotic leg will be integrated with sensing systems, upper-body support mechanisms, and human–robot interaction interfaces to form a medical-assistive robotic platform capable of performing walking-assistance and rehabilitation tasks. Functions such as assisted walking and stable support will be validated through an experimental platform, gradually improving the system’s applicability and reliability in practical care settings.
Innovation
The innovation of this study is mainly reflected in the integration of bionic leg mechanism design and human–robot interaction control, with a focus on medical walking-assistance needs.
First, unlike general humanoid robots that emphasize high degrees of freedom and human-like motion demonstrations, this study focuses on stable load-bearing and safe walking assistance. The proposed robotic leg system is designed to better meet the needs of medical and long-term care settings, thereby improving the robot’s ability to physically support patients and assist with mobility.
Second, this study adopts a crank-linkage-based bionic leg mechanism. Through mechanical optimization and torque amplification within the mechanism itself, the system can reduce its reliance on high-power motors and large reducers, while improving energy efficiency, load-bearing capability, and system safety.
Third, this study integrates gait generation with human–robot interaction control. As a result, the robotic leg can not only walk autonomously, but also provide assistive force according to the patient’s movement and contact state, further supporting rehabilitation training, walking assistance, and care-related tasks.
Fourth, this study emphasizes the integrated design of mechanisms, sensing, and control. Through joint force sensing, posture estimation, and impedance control, the robotic leg can perceive and regulate interaction forces with the human body, thereby enhancing the safety, comfort, and clinical application potential of medical-assistive robots.
Expected Outcomes
The expected outcome of this study is to complete a prototype bionic robotic leg system with stable load-bearing and walking-assistance capabilities, while establishing related technologies in mechanism design, dynamic analysis, gait planning, and force feedback control. Short-term outcomes include the design of the bionic leg mechanism, development of gait generation methods, and establishment of joint sensing and control platforms. In the mid term, the robotic leg will be integrated with the medical-assistive humanoid robot subsystem. In the long term, application scenarios such as assisted walking, obstacle crossing, and rehabilitation training are expected to be demonstrated.
In terms of academic impact, this study will advance research in bionic mechanism design, robotic walker control, physical human–robot interaction, and rehabilitation-assistive robotics. It can also serve as a foundation for future publications related to advanced robotic control and medical-assistive technologies.
In terms of practical impact, this study will contribute to the development of assistive robotic technologies that can support medical and long-term care settings. It is expected to reduce the physical burden on nurses and caregivers during patient lifting, mobility assistance, and rehabilitation support, while improving the safety of patients during walking training and daily mobility.
In terms of talent cultivation, this study integrates mechanical design, control engineering, sensing technology, robotics, and medical-assistive applications. It will provide students with interdisciplinary hands-on training and system integration experience, helping cultivate talent for future smart healthcare, intelligent care, and advanced robotics industries.