Improved Gait assistance via Interactions with Tactile interfaces – i-Gait

i-Gait aims at developing a new human-robot interaction approach based on the sense of touch, which is expected to help improve fall prevention in the near future. Falls are a major health, societal, and economic problem associated with multiple traumas and high morbidity. They result in 6000 fatalities per year in France alone. The most common mean to prevent falls is the use of the cane.

The recent years have seen the emergence of robotic canes. These devices are generally made of basic canes mounted on wheeled mobile robots. To prevent falls, robotic canes adapt to their users’ balance. For example, if balance is about to be lost, a robotic cane will stop rolling to provide physical support. However, if users do not apply the appropriate force or do not adjust their posture, balance remains threatened. One approach to overcome this consists in providing additional sensory information, e.g., audio signals or vibrations, about the force applied to the robotic cane or about weak balance. Nonetheless, in this approach, users should consciously interpret the sensory signals and decide how to change their posture and/or their interaction with the robotic cane. This could even worsen balance.

New research results lead us to consider another approach. Indeed, we have reported that an individual’s centre of pressure can be displaced precisely, without any conscious cooperation, through an appropriate control of the motion of a surface, which the user touches lightly with his/her forefinger. In the same time, findings from the haptics community, showing that mechanically stimulating the skin can yield perceived compliance when pushing against rigid materials, have been recently reported.

The ambition of i-Gait is to outreach the state-of-the-art of robotic walking assistive devices by initiating automatic (i.e. unconscious, involuntary) precise adjustments of the postural balance and of the force applied to the cane by mechanically stimulating the skin of the hand palm. To reach this ambition, a tactile handle, which delivers computer-generated tactile feedback able to initiate the necessary postural and force adjustments, will be developed. The tactile handle will be mounted on the top of a robotic cane. All the developments will be evaluated experimentally, and a special attention will be paid to human factors, ergonomics, and acceptability. The main scientific bottleneck consists in generating the appropriate tactile stimuli that can influence balance and the interaction force. The technical challenge consists in developing a mechanically robust hand-held tactile interface that generates rich tactile feedback.

The outcomes of the project are numerous. First, a new human-robot interaction approach, consisting in initiating motor actions by the stimulation of the tactile channel, will be proposed. Second, driving balance through the tactile stimulation of the hand palm is a new result, which may allow a better understanding of the multisensory processes underlying postural control. Finally, our results are likely to pave the road to a promising commercial product that may benefit from the growth of the Silver Economy, which will represent a market of 130 billions euros in 2020, in France.

This original project is based on the recent results of the principal investigator. To reach the project objectives and raise the scientific and technical challenges, a multidisciplinary group, which expertise ranges from motor control and perception to robotics, supports the principal investigator.