Enforcing Safety Requirements of a Knee Orthosis Using Finite State Machines

Given the critical role of lower limb orthoses in enhancing the mobility of individuals with gait impairments, these have to satisfy a set of requirements to ensure safety properties under various conditions. Such safety requirements could be related to properties of the orthosis such as support torque or operating time. Different works have been proposed to control lower limb orthoses based on their torque output. However, safety guarantees (e.g., a battery-operated orthosis should keep providing torque support for a guaranteed minimal time of use) have not been considered so far. In our work, we introduce a methodology for controlling a given set of requirements of knee orthoses. We show that by incorporating an additional control layer that allows the system to react to the violation of any given requirements of the orthosis by dynamically adjusting the configurations of the orthosis system. This control layer is formally modeled using finite state machines (FSMs) which enables the application of model-checking techniques to provide formal guarantees regarding the controller's ability to satisfy the given requirements. For illustration, we apply the proposed methods to a digital twin model of the user's leg with the knee orthosis. The digital twin incorporates biomechanical and kinematic elements essential for accurately simulating the orthosis system. Experimental results demonstrate the controller's ability to maintain different orthosis properties (i.e., torque and operating time) within desired bounds, consequently satisfying the given requirements of the orthosis.