Smart Dance Mirror AI Voice Agent
Project type
Personal project
Role
Conversational Designer, Interface Designer
Timeline
1 month (2023 Spring)
Tools
Voiceflow, Figma, Adobe Premiere Pro
Overview
Problem
We found that people trying to learn a dance on their own find it difficult to find the dance they want to learn and to break down the choreography from just pausing or slowing down the video.
Solution
To reduce the inefficiency of learning a dance from home without an instructor, we make use of both a conversational voice agent and a smart digital mirror to create a fun and easy learning experience.
Research
Methods:
5 contextual inquiries to interview participants about their dance learning experience
3 experiential research to observe how people learn to dance at home
1 survey to understand how their individual learning experience can be improved
We mapped out the current user journey map after the research to further emphasize with users and acknowledge their pain points in learning dance.
Insights
Problem Statement
How might we improve the experience of properly learning dance choreography in the convenience of people’s own homes?
Ideation
Brainstorming
We each brainstormed possible solutions that address the problem. The idea of including a conversational voice agent on a dance mirror came to our attention as it could be both a visual and auditory experience.
Competitive Analysis
We found inspiration from existing products such as the motion sensing function from Just Dance and the use of a screen in a mirror from MIRROR by Lululemon. Thus, we think our idea of dance lessons that use a conversational voice agent could be an add-on feature for existing products such as MIRROR.
Just Dance
MIRROR by Lululemon
Prototype
New User Journey
Choree will allow users to define their skill level and song choice while also making recommendations based on their previous interactions. From there, Choree will teach the user the selected dance through step-by-step instruction and continuous feedback on different aspects such as form and timing.
Conversational Voice Agent Design
I built the happy path script before moving on to Voiceflow and building any error and repair paths to account for different use cases.
Initial Happy Path
Watch this video that I edited to see how users can interact with the Choree! (est. 2 minutes)
Features covered in the video:
suggest dance based on user need (eg. song choice, difficulty level, genre)
scan user movements to provide real-time feedback
usage of conversation to go back and forth to practice certain moves/sections
smart record and save video
visual guidance of both AI teacher and reflection of your body
Interface Design
Visual Aids for Longer Songs
Users can refer to the top timeline bar to look at their progress in relation to the song length. They can refer to the bottom illustration to know which section they are learning.
Choose your own instructor
You can either choose the default AI character or your idol to learn from. The appearance of the AI teacher and voice will change based on your choice.
Change voice based on dance genre
The tone of the voice will change based on the genre of dance you choose.
Impact
Supporting more complex interactions
Gamification elements in the mirror, such as points or badges, to motivate users to continue learning and improving their skills
Connect with social media platforms, such as Instagram or TikTok, to create a community of dancers who can share their progress, provide feedback, and learn from each other
Virtual reality experiences to provide a more immersive and engaging learning experience for users
Value Creation
Takeaway
Designing an engaging experience goes much beyond visual
Leveraging existing products and services in my design
Experience prototyping can unveil a lot of problems that normal research methods don’t
I have mostly focused on the visual aspect throughout my product design experience. I learned that user experience can be greatly enhanced if multiple sensory systems are involved. The conversational design is definitely something that I want to explore further.
I learned that a lot of good ideas starts from scratch and a lot of them don’t. Some great ideas could be an add-on to the existing technology and products to maximize its value.
Even though we have done a lot of interviews to find user pain points, a lot more pain points that users encountered were discovered when we used experience prototyping to observe how users learn or teach them how to dance.