 This study proposes a novel algorithm which can accurately identify the type of walking pattern of a person based on their ankle movements. It uses a single du FIMU axis to extract kinematics data from the ankles and then classifies the walking pattern into upstairs, downstairs, treadmill, overground and stationary. The algorithm was tested on 40 healthy participants and achieved an accuracy of 89.57% and 87.55% in the left and right sensor, respectively. Additionally, the algorithm was able to detect inherent multimodal walking features in elderly people, similar to those seen in young adults. This suggests that the algorithm has the potential to be used in robotic gait intervention for multimodal walking. This article was authored by Yogesh Singh and Vinit Vashista. We are article.tv, links in the description below.