INPOSTURE

Transforming sedentary lifestyle

ACADEMIC · USER RESEARCH· UX DESIGN· DESKTOP APP
AI based posture detection and tracking desktop application that helps white collared professionals correct their posture while working for extended number of hours.
Vax cover image that consists of the latest design changes and a few components

Project Overview

Contribution

Formative research
Ideation
Storyboard
Prototyping
Interaction design
Usability Testing
Visual Design

Tools

Figma
Google teachable machines
Miro
This was an academic project as part of my Interaction Design Practices course at Luddy School of Informatics and Computing
As a product designer, I managed a team of 4 designers and oversaw the entire design process from user research to usability testing

Company

Project duration

Fall 2021 (6 weeks)

Role

Product Designer
Project Background
The technology boom in the past two decades has increased the number of sedentary jobs by 80% in the US. These jobs require people to be mostly sitting in front of a personal computer while working for long hours in poor posture.

This can have some short-term implications like back pain, sprains, and long-term implications such as a herniated disc, spinal dysfunction, and joint degeneration. Most people feel these short-term effects from time to time but do not realize their severity.

Musculoskeletal disorders are associated with high costs to employers such as absenteeism, lost productivity, and increased health care disability, and worker’s compensation costs.
Current User Journey - Fatigue vs. Recovery Curve
Poor posture for a long time can lead to work-related musculo-skeletal disorders (WMSD) such as back pain, spinal dysfunction, joint degeneration, rounded shoulders.
Problem Statement
How might we motivate white-collared professionals with sedentary lifestyle to maintain right posture and prevent long term health implications

Problem Overview

Lack of awareness about body posture while working for long hours

Not motivated enough to take action, though users are aware of long term health implications

Physical inactivity of sedentary jobs with longer sitting streaks with no frequent breaks

Non availability of the right resources

Solution
InPosture is an AI-based real-time posture detection and tracking desktop application that helps white-collared workers with sedentary jobs to maintain the right posture
How does this work ?
The application uses the existing camera of the user's personal computer to capture the real-time posture. Posenet is a real-time pose detection technique with which you can detect human beings' poses in image or video.
Key Features
01
Real-time posture status indicator UI element
A floating UI element that can be dragged and dropped anywhere on the screen which subtly reminds users to correct their posture without interrupting with their work.
People tend to make decisions unconsciously hence small cues or context changes in color and memoji emotions can encourage users to make a certain decision without forcing them.
User is sitting in right posture
As user ignores posture, green color starts fading from the top
Red color starts filling after 5 minutes of poor posture
After 10 minutes of poor posture
02
Alerts, Break reminders and activity recommendations
App alerts the user every 20 minutes if they ignore the posture indicator and also recommends activities in break time as a goal
03
Posture Analytics
User can have an overview of their posture health over a period of time
04
Earn incentives
Users can earn badges and coupons for maintaining right posture with timed goals
Diving deeper in the process
Formative Research
In the research phase, we conducted an extensive literature review of 20 journal papers, online articles, interviews, field study and a market research. To study was conducted to have a holistic perspective of the problem space.
Interviews
White Collared Professionals
White Collared Professionals with musculo-skeletal disorders
Orthopedic Surgeons
Physiotherapists
We interviewed subject matter experts and a lot of different roles within this domain to gain qualitative insights of why white-collared professionals tend to ignore their posture.
Observations
We conducted 60 minutes field study at multiple locations that helped us understand the behavior of target users in a natural setting. We took down the notes and depicted rough sketches to record our sessions. These observations helped us analyze the behavior pattern of the users in the working environment.
Market Research
We conducted competitive analysis and market research to identify the current design conventions and product features of existing posture correcting solutions
Yes
No
Others
Inadequate feedback, User ends up switching off the device
Elastic body belts can cause discomfort eventually weakens muscles
Active Solutions
Posture correcting sensors, and smart wearable devices
Examples
Yes
No
Poor and moderately inaccurate
Expensive
Alerts and reminders
Cost
Sensing capabilities
Moderately priced
Nil
Posture tracking
Passive Solutions
Ergonomic Chairs, Elastic body belts and Cushions
Analyzing data
Once we finished our data collection, we collaboratively conducted an affinity diagramming session to find emergent themes and patterns among the data. We also constructed an empathy map which helped the team empathize with our target users and understand their behaviors, motivations and frustrations
What were the key findings from our formative research?
Based on our identified themes and patterns we narrowed down our score to the following pain points:
01
Users are not conscious about their posture
White-collared professionals get too engrossed in their daily deliverables, not realizing their posture that might lead to back pain.
02
Sensory feedback over cognitive feedback
Users tend to tune out the stuff they get repeatedly exposed to such as alerts from mobile applications and smartwatches. They preferred sensory feedback such as haptic vibration to fix posture.
03
Physical inactivity of sedentary jobs
The human body has evolved to walk, run and jump rather than being seated for an extended number of hours at one place without frequent breaks.
04
Lack of awareness on posture implications
Early professionals are unaware of right posture, body ergonomics and long term implications of poor posture. Even if they're aware, they're not motivated enough to take action.
05
Remote work has aggravated WMSDs
The pandemic accelerated existing trends in remote work. Often, professionals get too comfortable working on non-ergonomic set up such as a couch or lying down on bed while at home.
06
Expert recommendation
Regular exercise and stretching, ergonomic furniture and paying attention to the way your body feels can make a difference in terms of long term posture goals.
Personas
We devised the following persona based on the data collected. Below persona reflect an overview of our target users.
Sara Jacob, 23
Software Engineer, New York
"I have a hectic schedule on weekdays and end up waking up with back pain the next morning. I think that it might be because of sitting all day in poor posture while working. I wish to find more about this"
Ideation
To explore the maximum number of opportunities, we conducted 3 brainstorming sessions generating at least 80 ideas. We performed an affinity mapping session to eliminate ideas based on key constraints, economic and technical feasibility. The idea of InPosture was an amalgamation of concepts and key features from these 3 brainstorming sessions.
Rationale behind the design
Requirement

-
Make users aware of their posture

- Take frequent breaks while working

- Motivate users to sit right posture

Constraints

- Solution must uphold user privacy

- Should not create cognitive overload

- Should not interfere with user's environment

- Economical
Features

- Real-time user posture detection using    existing camera

- Alerts, and break reminders

- Timely posture analytics

- Privacy toggle switch to enable or disable    posture tracking according to user's    convenience
Storyboard
I created storyboards for the 3 best possible design directions which helped us take the focus off our internal bias and empathize with our users. The following image depicts a storyboard of a white-collared professional using our product InPosture at the workplace.
Design
As a team, initially we divided tasks to create wireframes and collaboratively discussed the workflows to bring about the efficient flow of the overall product experience.
Design Exploration
We conducted multiple whiteboarding sessions to review each of our wireframes which allowed exploring varied approaches. This session helped us create a seamless experience that is well thought through with a minimal number of steps to complete a task.
Design Iterations
The app has gone through 3 iterations of design cycles, with the most recent one being the biggest update to its visual design and suite of features.
Iteration 1

The initial idea of the design was to use the user's computer camera LED to represent the user's posture. Showing the time-based intermediate state that represents the user's posture was not possible with camera LEDs.
Iteration 2

We created a flat UI element consisting of the product logo that starts filling up as the user starts to ignore their posture. We could achieve an intermediate state which helped users maintain the right posture.
Iteration 3

I wanted the UI element to be subtle and blend with the ecosystem. emojis/ memojis are the fastest way to convey a message, These memojis can create subtle context change along with background which can make users make decisions unconsciously.
Evaluation and Testing
We wanted to evaluate our product and reflect on design decisions. Hence, we conducted usability evaluation with potential users, Heuristic evaluation, and expert evaluation. The following are the results from our evaluations:
This is some text inside of a div block.
What worked well?

1. Participants had positive feedback towards the UI indicator for real-time posture status.

2. Users liked the flexibility to turn off the app using a toggle on the top navigation bar.

3. Users appreciated recommendations based on posture statistics.
What could have been done better?

1. Participants were concerned about the battery usage of the app for continuous monitoring of posture.

2. Users were concerned about camera hardware dedicated to the app most of the time.

3. The users felt the need for an application walkthrough to understand the UI element, analytics, rewards, and the privacy toggle switch.