A review on gait kinematics acquisition sensors and its advancements in IoT and machine learning

The increasing need for analyzing lower limb kinematics urges the application of Internet of Things (IoT) and machine learning for advancement in sensors, towards accurate, real-time, fast, and easy access measurements. Lower limb kinematics such as joint angles and motion trajectories are necessary to distinctly explain the functionality of the locomotion. In clinical gait analysis, the details of locomotor function are important to identify abnormalities such as cerebral palsy, hip osteoarthritis and stroke. In sports, lower limb kinematics are used to evaluate the performances of sports' athletes. On the other hand, robotic prosthesis devices use joint kinematics for movement control. Among the kinematics sensors, laboratory systems are the gold standard which attaches reflective markers on the wearer and capture movement using infrared camera systems. However, laboratory systems are expensive, time consuming, require experienced professionals for data recording, and limited in space which has a tendency to alter natural walking patterns. Furthermore, laboratory systems are not real time analyzed with IoT or advanced protocols. Applications of wearable kinematics sensors (e.g., inertial measurement units and force-sensitive resistor instrumented foot insoles), enabled with IoT are recognized as an emerging field for outdoor gait monitoring. Current research trends towards applications of machine learning to enhance and improve the quality of kinematics data acquisition and analysis of wearable sensors. As such, this chapter presents a detailed systematic review of kinematics systems (from traditional to the latest advancements), its potentials, limitations, and future trends especially in the IoT and machine learning field. A detailed understanding of available methodologies is important to investigate possible directions and areas to be improved.