top of page

Hover

Brief

A gesture-UI-driven speaker system designed to work with Spotify at home.

Duration

2 weeks

Class

Experience - Aynne Valencia

Teammate

Sarah Huo

Keywords

Home Entertainment,

Natural User Interface, 

Rapid Prototyping, Agile Design Methods

hifi.185 2.png
Secondary Research
Secondary Research

Due to unprecedented events taking place around the world, a lot of us may be finding ourselves staying home more than usual these days. The growth of working from home has produced permanent changes in the ways we use our time. And we’ll be needing things to watch, listen, read and do when we’re done working.

Use of Technology

During the past decade, companies from different industries have been studying NUI (Nature User Interface) to achieve a more natural and immersive interactive experience.

For example, BMW, Microsoft, and Sony are all researching and implementing gesture control to reinforce the experience of their entertainment products.

 

Gesture control applications can be used at home, like home control terminals or simply a piece of furniture. 

Competitive Landscape (Gesture Controlled Speaker)

Shorter Learning Curve

Bastron

823da762-f51f-4127-b06d-4d0e3cc2c713_1.0
osund_gesture.jpg

Osound

logo-hover-02.png
desk-patch-render.png

Less Function

SONY LF S50G

More Function

1.jpeg

SONY Onebox

61VSKs5DDGL._SL1200_.jpg

Longer Learning Curve

Observation & Potential Pain Points

  • Latency & feedback

  • User onboarding & education

  • Comfortableness

  • Interaction Zone

Challenges

  • There are no established standards.

  • It’s easy to fall prey to the trap of “default thinking”.

  • It’s vital to understand what types of gestures to support and how to introduce them properly.

User Research
User Research

User Interview

We interviewed people who use wireless speakers for Music Streaming Apps. Interviewees were made sure to be regular users of Spotify through screening questions.

Together we talked about 15 minutes with each participant, uncovering little bits of information about who they are and what their experiences are like of using Spotify and their speakers.

Below are the synthesized results (The more dots means more participants share the same behavior).

Screen Shot 2020-07-28 at 12.19.36 AM.pn

Interview Summary

Users care about

  • Quick and easy way to search for music

  • A huge library of musical content

  • Easy socialization with people outside the app

  • High-quality sound

Users get upset about

  • Intrusive Advertisement with a free account

  • Undesirable music feed

  • Unsmooth experience connecting with wireless devices

Users are motivated by

  • Weekly/Daily Mix

  • Curated / personalized playlists

  • New music right up to their taste

  • Music shared among friends

Users use the speakers while

Screen Shot 2020-07-27 at 12.54.07 AM.pn

User Types

We categorized them into three types of users based on their purpose.

The Seeker 

  • Want a steady stream of good music recommendation.

  • Daily Mix satisfies as long as they don’t repeat too much

  • Rely on Spotify for new music discovery

The Heavy Rotator

  • Dedicated to their favorite bands.

  • Maintain a selected library. 

  • Use Spotify to cue up the latest releases. 

  • Concerned about sound quality.

The Casual​​

  • Enjoy different genres  of music in different scenarios

  • Often play what the system suggests

  • Won't bother to build their playlists

Journey Map-01.png

Current User Journey Map (Non-Gesture-Controlled Speaker)

Concluding from their story and our own experience, we made this journey map to help inform our persona and product space.

Persona

The journey map helps us organize thoughts and patterns uncovered from our interviews into the creation of our user persona. This persona represents the target needs and aspirations of our user.

Screen Shot 2020-07-27 at 5.56.11 PM.png
Concepts
Design Opportunities
& Concepts

In order to turn our findings into actionable statements, we used a technique called How Might We statements. This allowed conversations around our problem statement, enabling rapid brainstorming sessions to generate as many ideas as possible. 

Once we had about 20-30 HMW statements, we narrowed down the top four.

Final Concept(For User Testing)

Screen%20Shot%202020-07-28%20at%2012.17_

Revised Storyboard

Screen%20Shot%202020-07-28%20at%2012.18_
Problem Statement
Problem Statement
- How might we provide users with a simple
and smooth sound experience that requires minimum attention?

- We do it by removing the disruption of having to physically interact with a touch screen or be in front of the device.

Target User

  • Music lovers who enjoy music surroundings at home.

  • Use speakers to listen to music at home

  • Use a specific type of music streaming app

  • Live alone or with roommates

  • Aged between 24 and 34

Scenario

When they are occupied with primary tasks, for example cooking, working or exercising

Painpoints

Users get upset about   

    1. Unreliable discover feature

    2. Having to interrupt their current activity to make adjustments

    3. Unfriendly or delayed feedback

So they cannot focus on their primary task or build highly personalized playlists, and have bad experience using the speaker with Spotify.

Goal

To provide users with a simple and smooth music experience that requires minimum attention

Solution

A speaker that can   

    1. Be operated anywhere without direct contact

    2. Help users quickly "like" or "dislike" songs in the playlist

    3. Remember playlists for different usage scenarios 

    4. Provide instant feedback without interrupting the music flow

Prototyping & Testing
Concept Test

Low-Fidelity Prototype

As this phase, we used the gesture guide of competitor products as reference and made sure they are borrowed from real life and feel natural.

unnamed (4).png

​Take off the sensor patch

unnamed (3).png

Stick the patch onto any surface

unnamed (2).png
unnamed (1).png

Basic Controls

(Play/Pause, Switch Song, Adjust Volume)

Personal Preference Controls

(Not interested, Like)

Screen Shot 2020-07-29 at 5.30.41 PM.png

Gesture Guide V1

 Concept Test

We showed the video to participants, walk them through our concepts and asked about their thoughts. Then the users were asked to make corresponding gestures by looking at the guide (without oral prompts)  and scored each gesture in terms of learnability and intuitiveness (From 1 to 5).

Test Result

While these results are disappointing, in that this concept has not been positively endorsed by the participants, neither has it been rejected. 

Our hypothesis was that the video failed to show the value of this solution to full potential, considering that the prototype was not realistic and we didn’t put the users in that context. And we needed to iterate and test more to improve our gesture set.

Gesture Iterations

Evaluation

When developing NUI, user testing takes a primary role. What is natural for one person, may be strange for another. Upon testing, the most natural option is the one that is preferred by the most amount of people that took part in testing.

After the first round of testing, we decided to take a step back and evaluate gestures with the four metrics below.

 

(Reliability refers to the likeliness of the hand tracking device correctly interpreting the gesture. Comfort refers to how physically difficult a gesture is to perform. )

Screen Shot 2022-10-10 at 7.07.12 PM.png

The raw data was bias because our sample wasn't including people with certain diseases or disability, for whom some gestures that involve rotating the wrist and finger poses can be hard to perform. So we down-graded these gestures based on existing study result. 

Test Round 2

We tested with a different group of participants here, and the scores are relatively higher than in the first round.

Screen Shot 2022-10-10 at 7.26.39 PM.png

Gesture Guide Iterations

Screen Shot 2020-07-30 at 2.30.05 PM.png
Screen Shot 2020-07-30 at 2.30.26 PM.png

Final Gesture Guide

Screen Shot 2020-07-30 at 7.48.36 PM.png
Screen Shot 2020-07-30 at 7.48.49 PM.png
Screen Shot 2020-07-30 at 7.48.36 PM.png
Mobile Touchpoint

Redesigned Spotify Flow

unnamed (8).png
unnamed (6).png

Add-on Device tab in Spotify enables user to manage playlists and preferences for the speaker

Assign playlists to  patches and store songs for different usage scenarios.

unnamed (7).png
unnamed (5).png

Users can easily access their favorite songs and personalized recommendations

A digital gesture guide to get users familiar with operations.

VisualLanguage
Visual Language

Brand Personality

moodboard-2-01.png

Quiet      Stable      Sustainable

Final Prototype
Final Prototype
intro 2.png

Gesture Control

​Want to control volume, pause or save music without being vocal or soiling the speaker? Just wave you hand over the top - no contact required.

hifi.185 2.png
hifi.182.png

Sensor Patches

​Users can place the sensor patch anywhere in the home, such as desktop, walls, bathroom etc. 

 

The four sensor patches have long battery life. Before you go to bed, you can put the sensor patch back on the backplate for re-charging.

Main Speaker

Magnetic Backplate

Press once to turn on the speaker

The light ring is on

When connecting Bluetooth

The light ring continuously blinks

When successfully connected

The light ring is off, sound is on

button.png

Power Button

patch-render 2.png

Visual Feedback

Breathing light rings on the sensor patches provides feedback without disturbing their music stream when user gives an order. The lighting effect applies to actions and states that do not have sound feedback (“like” ,“no interest” and everytime the speaker is turned on).

"Like" & "Not Interested"

The light ring blinks twice

"Play" & "Pause"

The ring gradually lights up or goes out

Running out of Battery

The light ring blinks in red

desk-patch-render.png

Breathing Light Ring

Anticipated Experience Map
journeyMap_Hover.png
bottom of page