More versatile, user-independent tools for recognizing and predicting locomotion modes (LMs) and LM transitions (LMTs) in natural gaits are still needed. This study tackles these challenges by proposing an automatic, user-independent recognition and prediction tool using easily wearable kinematic motion sensors for innovatively classifying several LMs (walking direction, level-ground walking, ascend and descend stairs, and ascend and descend ramps) and respective LMTs. We compared diverse state-of-the-art feature processing and dimensionality reduction methods and machine-learning classifiers to find an effective tool for recognition and prediction of LMs and LMTs. The comparison included kinematic patterns from 10 able-bodied subjects. The more accurate tools were achieved using min-max scaling [-1;1]interval and “mRMR plus forward selection” algorithm for feature normalization and dimensionality reduction, respectively, and Gaussian support vector machine classifier. The developed tool was accurate in the recognition (accuracy > 99% and > 96%) and prediction (accuracy > 99% and > 93%) of daily LMs and LMTs, respectively, using exclusively kinematic data. The use of kinematic data yielded an effective recognition and prediction tool, predicting the LMs and LMTs one-step-ahead. This timely prediction is relevant for assistive devices providing personalized assistance in daily scenarios. The kinematic data-based machine learning tool innovatively addresses several LMs and LMTs while allowing the user to self-select the leading limb to perform LMTs, ensuring a natural gait.

This paper presents a cost- and time-effective wearable inertial sensor system, the InertialLAB. It includes gyroscopes and accelerometers for the real-time monitoring of 3D-angular velocity and 3D-acceleration of up to six lower limbs and trunk segment and sagittal joint angle up to six joints. InertialLAB followed an open architecture with a low computational load to be executed by wearable processing units up to 200 Hz for fostering kinematic gait data to third-party systems, advancing similar commercial systems. For joint angle estimation, we developed a trigonometric method based on the segments’ orientation previously computed by fusion-based methods. The validation covered healthy gait patterns in varying speed and terrain (flat, ramp, and stairs) and including turns, extending the experiments approached in the literature. The benchmarking analysis to MVN BIOMECH reported that InertialLAB provides more reliable measures in stairs than in flat terrain and ramp. The joint angle time-series of InertialLAB showed good waveform similarity (>0.898) with MVN BIOMECH, resulting in high reliability and excellent validity. User-independent neural network regression models successfully minimized the drift errors observed in InertialLAB’s joint angles (NRMSE < 0.092). Further, users ranked InertialLAB as good in terms of usability. InertialLAB shows promise for daily kinematic gait analysis and real-time kinematic feedback for wearable third-party systems.

Current research suggests the emergent need to predict locomotion mode (LM) transitions to allow a natural and smooth response of lower limb active assistive devices such as prostheses and orthosis for daily life locomotion assistance. The purpose of this work is to develop an automatic, user-independent system using an environment-aware strategy to predict LM transitions. We applied an infrared laser system to measure the distance between the user and the terrain ahead. A three-layer decision tree with heuristic decision rules only dependent on infrared laser features was implemented to predict LM transitions. The prediction system was validated with 10 healthy subjects that performed 8 LM transitions in different terrains (level-terrain, stairs, and ramps). The results showed a prediction accuracy above 80% for all LM transitions, achieving 100% prediction success for transitions ramp/stair descend to level-terrain. All the LM transitions were predicted with high prediction time (> 0.73 seconds) which empowers its integration on assistive devices control strategies. The prediction system accurately and time-effectively predicts 8 different LM transitions only using the infrared laser sensor. It approached indoor and outdoor terrains, relevant for daily-life locomotion applications, and was more polyvalent and effective than previous environment-aware systems.

Neurological diseases may reduce Tibialis Anterior (TA) muscle recruitment capacity causing gait disorders, such as drop foot (DF). The majority of DF patients still retain excitable nerves and muscles which makes Functional Electrical Stimulation (FES) an adequate technique to restore lost mobility. Recent studies suggest the need for developing personalized and assist-as-needed control strategies for wearable FES in order to promote natural and functional movements while reducing the early onset of fatigue. This study contributes to a real-time implementation of a trajectory tracking FES control strategy for personalized DF correction. This strategy combines a feedforward Non-Linear Autoregressive Neural Network with Exogenous inputs (NARXNN) with a feedback PD controller. This control strategy advances with a user-specific TA muscle model achieved by the NARXNN’s ability to model dynamic systems relying on the foot angle and angular velocity as inputs. A closed-loop, fully wearable stimulation system was achieved using an ISTim stimulator and wearable inertial sensor for electrical stimulation and user’s kinematic gait sensing, respectively. Results showed that the NARXNN architecture with 2 hidden layers and 10 neurons provided the highest performance for modelling the kinematic behaviour of the TA muscle. The proposed trajectory tracking control revealed a low discrepancy between real and reference foot trajectories (goodness of fit = 77.87%) and time-effectiveness for correctly stimulating the TA muscle towards a natural gait and DF correction.