When life was collectively shut down by COVID-19, classrooms all over the world transitioned online. Teachers found themselves having to give lessons through video conferencing tools.
Captive8 is a video conferencing platform that leverages AI to help teachers provide better virtual lessons. Captive8 was designed during the 2020 University of Washington Husky AI Hackathon and was awarded 3rd prize out of ten teams.
Despite the time constraint and being the only UX practitioner in my team, I wanted some insight into the experiences of teachers who transitioned to giving lessons virtually. A quick search led me to a survey of more than 7,000 K-12 teachers in the USA. The findings from this survey guided our design direction.
https://www.washingtonpost.com/education/2020/04/27/teachers-parents-principals-tell-their-stories-about-remote-learning/1. Teachers experienced difficulty adjusting to teaching remotely.
2. Teachers were experiencing heightened levels of negative emotion.
3. Only a small number of students are engaged in virtual settings.
4. The majority of students were not expected to be on pace with learning
Feeling the pressure of the hackathon, I made the mistake of not carrying out proper ideation. Instead, we verbally brainstormed our solution. Through a quick discussion, we scoped our solution for K-12 teachers and came up with a design direction. Our hypothesis was that if we could assist them in keeping students engaged and making the transition more pleasant, we should see fewer negative emotions and improvements in student learning. Given that this was an AI hackathon, AI capabilities were considered in the final solution.
Captive8 is a video conferencing platform that helps teachers with student engagement. With Captive8, teachers are able to engage students attention by pushing tailored attention cues to distracted students. The AI learns which attention cues to send depending on students response to the different types of attention cues and individual subject performance. This makes the attention cues more personalized and better at drawing distracted students' attention back to the lesson.
The intention behind putting together a mood board was to identify colors that were present in classroom settings. The goal of this is to derive a color palette that would replicate the atmosphere of being in a physical classroom.
1. Captive8’s AI looks at inactivity or disappearance from the screen as indicators of distraction. When inattention is detected, the borders around the distracted student’s window lights up.
2. This signals to the teachers that this would be a good time to push an attention cue to the students. Attention cues are actions that students are prompted to perform. These can be key presses or short quizzes in the subject that the student performs best at.
3, Information on subject performance, responsiveness to different attention cues are stored. Teachers have access to this information in the student dashboard.
4. Teachers have autonomy over Captive8's AI where they can exercise discretion in deciding what attention cues to use depending on lessons outcomes and increased engagement.
Given the constraints of this project, I am proud of what was accomplished and the experience gained. Through this I learned to work in a team with cross-functional skills. I also gained invaluable experience advocating our design to a panel of AI experts.
If I had to do it again, I would have spent more time ideating and encouraging my team to contribute their own ideas. By encouraging individual ideation, we would have generated more potential design directions. This might have helped in addressing the teachers’ experiences in a more direct way.