In Accessibility taught by Professor Jeff Bigham, my team and I worked with a client with cerebral palsy to improve her AAC device system.

Observational Study

Non-verbal communication is a crucial aspect in everyday communication and face-to-face interactions. For people with complex communication needs (CCNs), augmentative and alternative communication (AAC) devices are designed to support and facilitate their conversations with others. One of the challenges with AAC systems, however, is their inability to portray components of nonverbal communication that typically accompany verbal speech.

I helped design an observational study with the purpose of gaining a better understanding of face-to-face interactions of AAC users and their caregivers, as well as existing strategies caregivers utilize to support AAC users in their communication with others. Furthermore, the results will provide insights on how to incorporate social cues and interactions during communication using AAC. These findings can inform the design of social agents and assistive technologies which support engagement and autonomy for AAC users with CCNs.

Case Study

Improving Usability of AAC Device for an Individual with CP

Apart from the observational study, my team and I worked directly with a client with cerebral palsy to help improve the design of her AAC device system. Individuals with cerebral palsy (CP) often experience difficulty speaking to an extent that threatens their ability to communicate (Pennington, Goldbart, & Marshall, 2004) due to motor skill deficits such as limited range of motion and/or lack of coordination (Koester & Arthanat, 2017). As such, these symptoms necessitate the use of AAC technologies that are designed to improve and maintain quality of life for individuals with complex communication needs.

Design Principles

We conducted research in physical therapy, joysticks, ergonomic handles, and other domains.

  • Obstacles of AAC devices

  • Physical constraints of AAC devices

  • The effect of joystick handle size and gain at two levels of required precision on performance and physical load on crane operators

Prototyping Process

Fig. 1: Our initial meeting with our client

Fig. 1: Our initial meeting with our client


Our client is a 37-year-old woman with CP. In our initial meeting, she addressed that her joystick was difficult to grip and control. To improve the interaction between our client and their AAC Device, we focused on the following goals:

  • Reduce fatigue when using device

  • Improve accuracy of selection on AAC device

  • Incorporate ergonomic design principles to provide long-term benefits

Fig. 2: Initial Prototyping

Fig. 2: Initial Prototyping

Initial brainstorming and solutions

The initial prototypes were 3D modeled using Autodesk Fusion, and printed using PLA material. During our first round of testing, we asked her questions about the dimensions, shape, and grip for each prototype.

Initial solutions to provide more comfort of use for our client included

  1. adding a platform that moved and elevated at an angle that would act as an armrest.

  2. moving the joystick and button configuration on the chair so the client would have easier access to the hardware.

  3. altering the location of the keyboard option on the AAC device that the client uses to reduce the distance of character selection on the device.

User Testing

During the next round of testing, we sought to operationalize the initial goals of improving usability, reducing fatigue, and testing durability of our prototypes, and came up with a list of metrics we wanted to measure during our observations. For each metric, we asked her to rate her experiences with the use of a Likert scale.   

We recorded:

    • Time (Timer)

    • 1-10, How well does the shape of the prototype fit? (1 being least fit)

    • 1-10, How tiring was this typing experience? (1 being least tiring, 10 most tiring)

    • 1-10, How well can you grip this prototype? (1 being worst grip, 10 best grip)

    • 1-10, How much do you like this prototype? (1 being dislike very much, 10 like very much)

    • Is the prototype too thin, too wide, or just right?

    • Is the prototype too short, too tall, or just right?

    • General comments from client

Our goal was to compare performance typing sentences of variable difficulty with each prototype. We asked the client to type sentences of variable difficulty, and recorded responses to questions about the prototypes after typing the sentences. Our baseline comparison was with the client’s original joystick.

We came up with three sample sentences for her to type.

Easy: The sky is blue.

Medium: When I go to music camp, I get to see my friends.

Hard: When you go to the store, could you buy some milk, cereal, and bread?

During our actual observations, we started with the easy sentence and went through each prototype. Due to limited time and fatigue effects, we had her type up a sentence and we followed up with the questions we had prepared in order to get her opinion on the models.

Fig. 3: Design with client’s measurements from our prototyping session results

Fig. 3: Design with client’s measurements from our prototyping session results

Solution Design

After we conducted our user testing, we found that our client preferred the cylindrical model over the others, rating it with the best fit, least fatigue, and best grip. In order to increase the durability of the final prototype, we 3D printed it so that it was not hollow, and therefore could withstand her strong grip. In future iterations, we plan to fit our joystick with a rubber or silicone coating to make it more comfortable for the client to grip.