Over a multi-week process, each lesson built upon the previous as the teacher, coach, and students all mastered the AI tools. After a few weeks, Bird and Ferris decided it was time to make their AI learning more meaningful through project-based learning. They looked for a problem within their students’ world.
As students of the ECHO Joint Agreement – EASE Program, Bird’s students were all served by Individualized Education Plans. Each plan is different because each student in Bird’s class may have unique supports needed. Students in Bird’s class have complex disabilities, including cognitive, language, physical, health and/or vision needs. Though each plan is individualized, one common thread among them was many received services from support staff, including physical therapy, speech and occupational therapy.
This sparked a powerful question:
What if we could use AI to help students practice their OT exercises independently when their therapist wasn’t available?
“It was a win-win,” Ferris remembers. “The support staff wanted the students to practice more than their short session time allowed, so this was a great way to develop something to help them.”
With the support of Bird and the support staff, the students started training the AI. They learned through perseverance and grit; the recording didn’t always work, they stood too close, or their poses weren’t quite right. With each challenge, the students adjusted their data, training the machine to work exactly as they intended.
“Students worked collaboratively to test and improve each other’s models through Google Classroom,” Ferris said. “They went through probably five or six different iterations. The students were thrilled every time we’d work on the project.”