Self-organization of the communication space based on user range-of-motion: a framework for configuring non-contact augmentative communication devices

Abstract

Augmentative communication devices are traditionally used by individuals with chronic communication disabilities. Recently, there has been a growing interest in deploying augmentative communication technologies when individuals temporarily and suddenly lose the ability to speak, due for example, to surgery, mechanical ventilation or disease advancement. Due to the complex and evolving physical needs in these settings, flexible noncontact communication devices are proposed as potential communication solutions. Non-contact communication devices typically translate some form of intentional movement of a user into an interface navigation command (e.g. mouse emulation) or directly into a target message (e.g. yes or no). Even with a modest vocabulary, optimally configuring a user's communication space by hand can be a timeconsuming process that needs to be repeated when the user's vocabulary needs or physical abilities change. We propose a general methodology for computer-vision based communication devices whereby the communication space can be configured automatically through mathematical optimization. The objectives of such optimization are to maximally exploit the reachable space, to minimize probability of user error and to minimize the user's average effort in reaching given target messages. In addition, selforganization of the communication space observes constraints reflecting the user's range of movement. We present a simple realization of this general methodology and exemplify some results obtained using a sample of a communication space.