STAR:T is an AI-powered dance coaching app for international K-Pop fans. Initially designed with XR technology, user research led to a pivot to AR, allowing for more practical and immersive learning. The app provides real-time feedback, multilingual support, and smartwatch integration, enabling seamless interaction with an AI coach.
Background
With the global rise of K-Pop, fans seek structured dance training, but platforms like YouTube lack personalized feedback. Our initial idea was to use XR for a fully immersive experience, but user testing revealed that VR hindered movement. This insight led us to adopt AR, ensuring dancers maintain full mobility while enhancing interactivity.
Research Insights
Through surveys and interviews with over 100 K-Pop fans, we identified:
Limited Feedback
86% of respondents rely on video tutorials, which do not offer real-time corrections or personalized guidance.
Device Handling Difficulties
Many users find it cumbersome to manage their mobile devices while dancing, disrupting the flow of practice sessions.
Language Barriers
Existing resources often lack multilingual support, hindering non-native speakers from fully engaging with the content.
VR Usability Concerns
Wearing VR headsets was impractical for dance, leading to discomfort and limited visibility.
Pivot to AR
User feedback revealed that VR technology's limitations—such as discomfort and reduced situational awareness—made it unsuitable for dynamic activities like dancing. This insight led us to pivot to AR technology, which allowed users to practice in their real environment while benefiting from interactive guidance.
Hypothesis & Goals
We hypothesized that an AI-powered app, complemented by smartwatch integration and voice recognition, would enhance the dance learning experience by providing:
Real-Time AI Feedback
Immediate corrections and guidance to improve dance accuracy.
Hands-Free Interaction
Seamless control via smartwatches and voice commands, eliminating the need to handle devices during practice.
Multilingual Support
Accessibility for a diverse user base through multiple language options.
Community Engagement
Features that foster connections among users, such as dance challenges and local meetups.
Solution & Key Features
AR Dance Guidance & Real-Time AI Feedback
Hands-Free Interaction
Multilingual & Community Features
Competitor Analysis
We evaluated existing dance learning platforms and identified a lack of interactive feedback and hands-free control options. This informed our decision to integrate AI technology with wearable devices and voice recognition to address these gaps.
We also assessed other dance learning platforms and identified gaps in real-time feedback and hands-free functionality, justifying our pivot to AR technology.
Design & Prototyping
Our design process included:
Wireframing & User Flow Mapping
Developed paths for AR interactions and smartwatch navigation.
High-Fidelity Prototypes
Creating detailed prototypes to test the integration of AI feedback, smartwatch controls, and voice interactions.
Information Architecture & User Flow
The app's structure is designed to support various user journeys:
Lesson Selection
Users choose from a range of dance routines categorized by difficulty and style.
Practice Mode
Engaging in practice sessions with real-time AI feedback, controlled via smartwatch or voice commands.
Performance Review
Analyzing performance metrics and receiving personalized improvement suggestions.
Community Engagement
Participating in challenges, sharing progress, and connecting with fellow learners.
Brand Identity
The name "STAR:T" combines "Star" and "Training," reflecting the app's mission to help users achieve star-level performance. The visual design incorporates vibrant colors and dynamic graphics inspired by K-Pop aesthetics, creating an energetic and engaging interface.
Smartwatch & Voice Interface Design
The smartwatch interface features intuitive controls for playback and feedback navigation, while the voice interface recognizes natural language commands, ensuring smooth interaction during practice.
Community Features
The app's community features are designed to foster user engagement and interaction:
Dance Battle Interface
A dynamic layout that allows users to participate in dance battles, view competitors, and track progress.
Group Challenge Page
A collaborative space where users can join group challenges, monitor team performance, and celebrate achievements.
Local Event Listings
An organized directory of local dance events, enabling users to discover and participate in community activities.
These GUI elements are crafted to provide an intuitive and immersive experience, encouraging users to connect and grow within the STAR:T community.
HOME : Class : Let's receive personalized K-POP dance coaching through classes!
Compare your movements with the AI instructor's on a large screen using screen mirroring
Participate in dance battles by matching with opponents ranging from friends to AI or random users. Communication through chat and cheering is possible during the dance battle
After the dance battle, receive detailed feedback on accuracy, identifying both well-executed and less accurate moves
Development Challenges
Despite extensive efforts to bring the app to market, a significant challenge was finding a developer skilled in implementing the motion capture feature—a critical element for real-time feedback. Although the app's primary features were developed, the absence of this functionality prevented a full launch. This limitation also meant we could not gather direct user feedback post-launch, which was a key objective of the project.
Challenges & Learnings
Collaborating with multiple designers on STAR:T was a valuable first experience. Without a designated leader, we relied on each other’s strengths. One teammate, focused on UX, frequently suggested visual revisions based on other teams' work, despite our visuals being aligned with user insights. To address this, I referenced interview findings to support our design choices.
This experience taught me the importance of data-backed communication and the need to balance personal connections with project goals, fostering adaptability and clear dialogue in team settings.
Outcomes & Recognition
The project received notable recognition, winning 1st place in the 2023 Day One Company App Service Idea Contest for its innovative approach. However, the challenge of implementing motion capture technology highlighted the importance of technical feasibility in project execution.
Reflection
This project emphasized the importance of user-centered design, adaptability, and the need to align technical capabilities with user needs. While we were unable to release the app due to the motion capture challenge, the experience provided valuable insights into the complexities of developing tech-driven solutions. Future steps could include partnering with specialists to incorporate the motion capture feature and gathering post-launch user feedback.
