• rehabilitation engineering
• assistive strategies
• gait phase estimation
• adaptive oscillators
• segmentation

Life expectancy has been increasing along with the growth of the industrialization. In order to improve the quality of life of elderly, many research teams focused on the development of new technologies, such as wearable robots, to help people remain productive and enhancing their mobility, thus preventing them from being affect by senile disorders and therefore reducing costs of the healthcare system. Different kinds of active powered orthosis were conceived in the last decades; nevertheless, the design of an effective and robust control architecture, that is adaptive to the wearer movement intentions is still a challenging issue.
The goal of this work was the development of an adaptive assistive strategy for controlling an active pelvis orthosis conceived for assisting hip flexion-extension during walking by means of capacitive sensors. The development of the proposed strategy resulted from three main activities: (i) an accurate off-line data analysis to identify the most suitable placement of the capacitive sensors on the lower-limb segments; (ii) off-line validation of the performance of a gait-phase estimator based on adaptive oscillators coupled with the capacitive signals; and (iii) on-line testing of the conceived control architecture for providing an assistive action proportional to the capacitive signal variation dependent on the gait-phase during walking. The effectiveness of the provided assistance was verified comparing lower-limb EMG activity during assisted and unassisted walking.