BIO-SYNC
ROLE
UX DESIGNER
CATEGORY
AR & MULTIMODAL
DURATION
FEB - APRIL
TEAM
SOLO PROJECT
Information Architecture
OVERVIEW
BioSync is an AR-powered multimodal biomechanics interpretation app that combines different sensing inputs to create meaningful understanding of one’s movement. The project focuses on helping people with MSK condition to track their movement and help them prevent future injuries by detecting their movements by replacing heavy medical jargon with intuitive visual guidance. This project applied the Lean UX method.
- Think: user interviews, literature review, affinity mapping, proto personas, and assumption prioritisation
-Make: hypothesis formulation, lo-fi wireframes, hi-fi prototype, and visual design system
-Check: three rounds of usability testing, SUS scoring, and hypothesis validation fed back into iteration
Key-Insights
Initial concepts focused on using high-quality IMU sensors and collecting data to give feedack.
However, research revealed that the core issue was not technology preference or the feedback, but medical jargon used for the feedback and no recommendations which lead to cognitive overload.
Patients didn’t need better technology. They needed a way to successfully understand their body and movement.
Research & Problem Definition
First-year students often struggle to navigate unfamiliar campus environments, especially under stress. This results in wasted time, increased anxiety, and missed academic activities.
Traditional maps become difficult to interpret
Students make repeated navigation failures
Many eventually give up and abandon their journey
I followed a Lean UX approach to validate assumptions and iterate on solutions quickly. This allowed rapid identification of high-impact usability issues early in the process.
Conducted 5 user interviews with university students
Synthesised insights using affinity mapping
Transformed assumptions into testable hypotheses
Iterated through lo-fi and hi-fi prototyping
The solution was tested through iterative usability testing.
Conducted 2 rounds of lo-fi testing to identify critical usability issues
Refined navigation entry point based on 0% task success in early testing
Validated final design with 5 users using a hi-fi prototype
Results:
100% task completion for core navigation flow
SUS score: 85.5 / 100 (Excellent usability)
Significant improvement in feature discoverability and task efficiency
Hypotheses Validation
Early lo-fi testing revealed a critical issue: users were unable to locate the warning, resulting in a 0% task success rate. This showed that the core feature lacked visibility despite its strong concept.
Based on this, I repositioned the warnings as an overlay and improved its visual clarity. I also refined UI affordances to clearly distinguish interactive elements. These iterations significantly improved feature discoverability and overall task flow.
Future Improvements
The most critical learning was the importance of challenging initial assumptions. The project began with a focus on emotional space recommendations, but research revealed that functional navigation failure was the real problem. This early pivot significantly improved the relevance and effectiveness of the final solution. Additionally, lo-fi testing proved essential in uncovering major usability issues before development, allowing efficient iteration with minimal cost.
AI/GenAI
I used AI tools at specific points in this project, deliberately and with full critical oversight. Claude assisted with structuring the interview plan, sourcing relevant research, and organising frameworks for usability testing. It also helped consolidate findings across test sessions. Grammarly was used for proofreading throughout the report. Gemini generated the proto persona illustrations. Every output was reviewed, questioned, and revised before it shaped any design decision.
1. Real time, Visual Gait Feedback & Simple Gait Score
Instead of raw sensor data, Users get intuitive, at-a-glance "gait score" and real-time visual feedback shown on the homepage.
Reduces cognitive load during stressful situations by making movement data understandable to non-clinical users the moment they need it.
Eliminates the fragmented, expert level output produced by current tools
2. AI Powered Insights and Coaching
AI interprets user's movements and give personalised recommendations in plain language
Addresses the comprehension gap
Easy to take action on
Early detection of harmful movement outside clinical environments
Reduces stress of doctor's visits
3. Post Session Summary and Progress Tracking
Users can track their progress and also access session summaries
Visible progress keeps the users motivated
Helps users mentally prepare their journey
Regular summaries create consistent engagement outside clinical visits.
THINK
SOLUTION
APPROACH
TESTING & ITERAIONS
WHAT I LEARNED
Design Process
Usability Testing
Solution 01
Solution 02
what participants have said about existing technology and their personal experience.
Problem
Solution
Confused "AI Insight" label as an interactive button rather than a status indicator.
Access AI Insight directly on the homepage opening it into an interactive chat
Problem
Solution
Made the warning pop up during the session and added a background sound to grab the attention
Nobody could locate the warning on the session page
Hypothesis 03
TARGET
Result
Improve Real-Time Gait Score
80%+
100%
Hypothesis 03
Reduce cognitive stress
TARGET
Result
90%+
100%
Hypothesis 03
Reduce Confusion
TARGET
Result
90%+
100%
Ghantt Chart
