We developed a pair of walker handles with haptic sensing capabilities, and a pair of walker handles with haptic feedback capabilities. Both pairs of haptic handles were parametrically designed to fit different walker frames and 3D printed using a conventional fused deposition printer (FDM).
Haptic Sensing Handles
The walker is a four-legged walker with custom haptic handles. The haptic handles contain 5 force-sensing-resistors per handle. Ergonomic curves are added to the handle to help ensure the user's hand is placed correctly over the sensors. In order to measure anterior and posterior forces, the main portion of the handle can freely slide over the walker's frame and is held in place between two rings, which are rigidly clamped to the handle. The handles are designed such that they can be added to any existing walker. These handles are used to measure information regarding physical interactions between the user and the walker.
These handles were used to collect data over the course of repeated demonstrations of correctly and incorrectly performed sit-to-stand transitions. This data was then used to classify the transitions as correct or incorrect. The classification method trains one Hidden Markov Model using the correctly performed transitions and one Hidden Markov Model using the incorrectly performed transitions. The resulting models are used to classify previously-unseen sit-to-stand transitions as either `correct' or `incorrect'. This classification proved to be highly accurate when classifying sit-to-stand transitions performed by two members of the lab.
The haptic sensing handles were also used to obtain gait parameters from healthy users. The gait parameters obtained from the walkers sensors were compared to gait parameters obtained using a motion capture system (OptiTrack).
Haptic Feedback Handles
This walker is a two-wheeled walker with custom haptic handles. The haptic handles contain two haptic feedback modalities embedded on them, vibrotactile feedback and skin-stretch-and-slip feedback. Two older adults, aged 66 and 76, participated in a pilot study to evaluate the perceptibility of the haptic cues, both while standing still and while walking over a smooth surface. Measures of the participants' ability to feel the haptic cues and their responses to questionnaires indicate that skin-stretch-and-slip cues are superior to vibrotactile cues for a smart walker to be used by older adults.
Cruz, Ramón E. Sánchez, and Rebecca P. Khurshid. "Design and Preliminary Assessment of a Haptic Feedback System For a Smart Walker." 2021 IEEE World Haptics Conference (WHC). IEEE, 2021.