D2: AI Augmented Schedule Management
Researchers and Product Developers are welcome to use our results!
- A video illustrating the application
- A paper with technical and evaluation details -
Edmund H. Durfee, Lynn H. Garrett, and Abigail Johnson. “Promoting Independence with a Schedule Management Assistant that Anticipates Disruptions,” Journal of Healthcare Informatics Research 4:19-49, 2020.
- A repository with our prototype source code
- A data repository containing a transcript of our final expert focus group session and a summary of the key themes identified
Principal Investigator: Dr. Edmund Durfee
Coordinator: Mrs. Donna Omichinski
Graduate Students: Ms. Lynn Garrett, Dr. Jason Sleight, & Mr. Drew Davis
Undergraduate Students: Mr. Jordan McKay & Mr. Drew Canada
Participation in everyday unscripted and informal life situations generally involves risk that the environmental and social context might differ from expectations. Physical and/or cognitive impairments typically accentuate such risks by reducing a person’s ability to respond to the unexpected.
In particular, adolescents and young adults taking on increased self-management responsibilities will typically lack the life experience needed for anticipating, recognizing, and overcoming such obstacles.
In response, they may overestimate the likelihood, severity, and/or inflexibility of such situations, while underestimating their competence for handling them, and thus unnecessarily miss out on enriching experiences within their grasp. Or they may underestimate the risks and overestimate their own abilities, and thus may put themselves in positions that leave them feeling physically or socially inadequate, and even risk their health.
This development project designed, prototyped, and evaluated decision-support technologies that augment cognition of inexperienced decision makers to avoid overestimating or underestimating risk so as to maximize safe, positive participation in meaningful activities and life experiences. The resulting application maximizes flexibility to the adolescent by offering a choice of any activities that could be done without violating constraints on getting everything done at appropriate times. The application helps visualize the downstream consequences of earlier choices in the day, and allows the user to add, remove, or modify activities as the day progresses.
Importantly for an adolescent transitioning to self-management, the application supports connections between the schedules of multiple people, such that, for example, decisions by an adolescent and a parent are reflected in each other’s schedules, allowing dynamic coordination over familial activities (e.g., dinner together) and healthcare activities (e.g., therapy that requires a parent’s help).
Over the project lifetime, prototype versions of the application were presented to focus groups that could include adolescent/parent dyads as well as clinical experts. By incorporating feedback, the prototype addresses needs that are not met in existing products, especially for adolescents with complex health challenges.
With the conclusion of the project, the concepts, technical insights, and evaluations of the developed technologies are now available for other researchers to build off of, and in particular for developers of commercial products to use to increase the utility of their products for all users, but especially for inexperienced users for whom errors in schedule management can have significant health consequences.