Video Demonstration of the Phase I Prototype: http://youtu.be/0FtoN0AX6PE

Purpose: This project team will develop and test an app, Speak Agent AAC, intended to increase communication rates and provide individualized supports to students with speech disabilities who use assistive technology to communicate. Among school-aged children with speech communication disabilities, students with Autism Spectrum Disorder are most frequently affected, as are students with cerebral palsy and apraxia. To communicate effectively, many of these individuals rely on Augmentative and Alternative Communication (AAC) systems, which convert symbols representing words and phrases into sentences read aloud by a computerized voice. While AAC systems do provide a means of communication, the rate (or speed) of current systems to convert symbols to computerized speech is often too slow for use in school settings. Furthermore, these systems do not offer customized supports to address the needs of individual users.

Project Activities: During Phase I (completed in 2014), the team developed a prototype app for students who require assistive technology to communicate. Components of the prototype included pictures of over 650 symbols that represent words, an assembly bar with words and phrases that students use to construct sentences, a text-to-speech engine, and a relational database used to convert the symbols to words. At the end of Phase I, pilot research with eight grade school students with speech disabilities showed that the prototype functioned as intended and that students were able to successfully generate sentences that were read aloud by the app. In Phase II, the team will add components to the prototype and build out the functionality of the predictive algorithm to customize the support the app can provide. After the development is complete, researchers will use a quasi-experimental research design (specifically, an ABAB reversal design) with approximately 16 kindergarten through grade 8 students diagnosed with Autism to test the effects of the app. In the study, groups of students will alternate using and not using the app over a series of weeks. Analyses will compare student communication rates and accuracy with and without the app during this period.

Product: Speak Agent AAC will be an app for tablets or handheld phones designed to increase communication and accuracy rates for students with speech disabilities. Students will use the app to construct phrases or sentences by selecting symbols, with the app converting input into computerized speech to enable effective communication. The app will generate customized profiles for each user by feeding data into its predictive algorithm, resulting in faster communication rates with greater accuracy for individual students based on their own usage patterns over time. The app will also include assessment diagnostics that service providers will use to guide instructional practice related to communication and speech.