Modeling an Automatic Adjustment of Scanning Rate Using Fuzzy Inference Logic

Single-switch scanning is one of the slowest way for people with severe motor impairments to access a wide range of assistive technologies, such as entering text into computer applications, manipulating alternative and augmentative communication devices, and operating power wheelchairs. Automating scanning rate adjustment has been investigated as a means of improving selection rate by several research groups [1-6], using probabilistic methods such as uncertainty factors and Bayesian networks. In practice, however, when assessing and documenting their client’s needs, clinicians tend to prefer specifying linguistic values, such as slow, medium, and fast, rather than numerical values. In this respect, fuzzy logic can be an alternative to the existing approaches because it can capture the meaning of linguistic variables whose values are words rather than numbers [7, 8]. This paper describes our approach to modeling the automatic adjustment of scan rate using fuzzy logic.