Marlin: A Wearable Coach for Swimmers

Best Master's project for MS-HCI, Runner-up
September 2016 - April 2017

Problem formulation through surveys, focus groups, and contextual inquiry

  • Users
  • Context
  • Problem
  • Swim Coaches
  • Training
  • Communicating the workout with multiple swimmers during training
    Monitoring their performance and communicating guidance
  • Swimmers
  • Training
  • Communicating with Coaches during training and keeping track of their performance

Insights Summary through competitive analysis and literature review

  1. Needs one-way communication between coach and swimmer
  2. Opportunity to track several metrics at once - Heart rate, Speed, Distance, Time, Breathing rate
  3. A solid modality to convey information to swimmers underwater is audio through bone conduction headphones
Ideation
With research, I identified that sonification of data is a viable solution to these problems. So, I decided that the coaches would have an app that allows them to plan a training program, and track individual swimmer's performance. The coach can operate this interface to set targets for each swimmer, and these instructions will be provided to the swimmer through sonification (headphones).

The swimmer, on the other hand will wear a wearable that tracks their heart rate, distance, speed, breathing rate, and time. This wearable will also provide them real time audio feedback, designed by a sound designer, about their metrics through bone conduction technology.
Prototyping
An early high level prototype of the app incorporates breathing rate, and the ability to set target values for each swimmer. But, after testing this with a few experts, I realised that planning a workout and setting targets requires much more customizability. Also, adding a direct functionality to speak to the swimmers using the iPad mic would be very useful.
As for sound design, I gave the sound designer two sets of calues for the heart rate and the stroke rate. The sound designer prototyped two versions of different timbres for the stroke rate., for each set of values. For both versions, the tempo of the heart rate varied with the values. As for the stroke rate, we decided to play one beat at the end of every stroke, like a reverse metronome. We also provided feedback about the time elapsed with a spearcon (speech icon).
Usability Testing
Ambushed a bunch of swim coaches and swimmers during a collegiate swim-meet at Georgia Tech
The app and the sound design were evaluated with competitive collegiate coaches and competitive collegiate swimmers with a System Usability Score (SUS). The aim of the evaluation was to see if this is a viable solution for their problem, and if it would be useful in their training sessions. The two versions of the sound design were compared by both the coaches and the swimmers to see which design was more understandable and clear.
STUDY PROTOCOL
  • Participants: 4 Coaches | Sprint and Distance | 20-30 | 3-10 years of experience

    Protocol:
    Background Survey
    Introduction to the system
    Demo of the app
    SUS Rating for app
    Sonification - Listening and Learning the audio prototypes
    Comparison between the 2 audio prototypes
    SUS Rating for audio
    Semi Structured Interview - Are the prototypes adequate in function, customizability and usability?
  • Participants: 3 Swimmers | Sprint and Distance | 18-25 | 3-15 years of experience

    Protocol:
    Background Survey
    Introduction to the system
    Sonification - Listening and Learning the audio prototypes
    Swim and Explore the 2 prototypes
    Comparison between the 2 audio prototypes
    SUS Rating
    Semi Structured Interview - Is the prototype clear and understandable?
RESULTS
Overall, the system was quite well received by all the participants in the study, and they did think that this was a usable solution to their problems while training.

SUS Rating for App: 68.1 (Good) | Tracker: 69.2 (Good)