Yarnkey: A Thigh-worn Wearable Input Device for 2D Alphabetic Patterns using Conductive-threads-based Touch Sensing
Ken Shibata (William Lyon Mackenzie Collegiate Institute)
Koya Narumi (University of Tokyo)
Overview
2
Background: Existing Wearable Keyboards
3
TelemetRing[1]
TipText[2]
Half-keyboard[3]
Wearable Clavier[4]
1D keyboard using conductive threads[5]
Background: Comparisons
4
Name | Method | Advantages | Disadvantages |
TelemetRing[1] (2020) | Rings and bracelet-type devices Using finger tap combinations | Lightweight | Multiple devices required to function |
TipText[2] (2019） | Finger sack type device Using finger tap combinations | Lightweight Small Similar layout of keys | Covers the tips of fingers |
Half keyboard[3] (2012) | Assign the half keyboard to the full keyboard for input | Similar layout of keys | Bulky |
1D Keyboard using conductive threads[5] (2021) | 1D handwriting character input device for wearable devices using conductive threads | Lightweight Thin Easy to learn | Harder to recognize some characters |
Project Jacquard Trucker Jacket[12] | Simple gesture input using incorporated conductive threads | Not visually invasive | Uses both hands/arms |
Design Goals of Yarnkey
□ Use of one finger
□ Non-visually-invasive, lightweight, and small
□ Allows multiple postures (i.e. seated and standing)
□ Easy-to-learn
5
Development Process
6
Normal keyboards require the use of both hands…
The upper thigh area where the hand naturally falls is the perfect placement, but a full keyboard requires too many sensors…
Let's write each letter as a single stroke that resembles the alphabet!
System Overview
7
Yarnkey is a wearable input device using conductive threads
a
b
c
d
e
f
g
h
i
j
Key map example
6-sensing-point data processed by microprocessor
Uses custom code[6]
System: Resistive Touch Sensor Hardware
8
8
R1
A/D
C
R1
R2
0V
VDD: 3.3V
A/D
VDD: 3.3V
A
B
A
B
(A) No contact
(B) In contact with finger
The A/D input voltage is provided through the pull-up resistor R1: �V1 = VDD = 3.3V
The A/D input voltage drops from VDD when touched by the ratio determined by R1 (on-board resistor) and R2 (human resistance)
👆
Prototypes
9
Using resistor-based touch-sensing, the sensor surface area is reduced to approximately 2 mm2 and therefore not visually invasive.
Keystroke Table
10
a
b
c d e
klmn
o
pqrst
uvwxy
space
enter
backspace
period
f g
h
i
j
z
�(https://nyiyui.ca/render)
Demo: Smartphone (https://youtu.be/ijQGRG3Er1g)
11
1
2
3
4
Demo: Smartwatch
12
1
2
3
4
Design Goals of Yarnkey
🗹 Use of one finger
🗹 Non-visually-invasive, lightweight, and small
□ Allows multiple postures (i.e. seated and standing)
□ Easy-to-learn
13
User Test Details
14
# | Name | Description | Time [mins] |
1 | Introduction | An introduction on how to use Yarnkey. | 5 |
2 | Learning Session #1 | Please use Yarnkey (with assistance) to become familiar with it. | 15 |
3 | User Test #1 | Please enter 5 sentences (without assistance, while seated). This will measure the number of characters entered and error rate within a specified time frame. | 5 |
4 | Learning Session #2 | Same as Learning Session #1 (step 2). | 15 |
5 | User Test #2 | Same as User Test #1 (step 3). | 5 |
6 | User Test #3 | Same as User Test #1 (step 3), but while standing. | 10 |
7 | Survey | Please complete the survey. | 5 |
App for User Test
15
User testing web app (https://yarnkey.nyiyui.ca)
Input Speed
16
(standing)
(seated)
Average for grade 2 students[9]
Test #2: The input speed was on average higher than grade 2 students.
Test #3: This test was conducted while standing.
Conclusion: Yarnkey can be used at around 0.5 characters / second while seated.
(seated)
Error Rate
17
(standing)
(seated)
Test #2: The error rate decreased by 3% after the 15-minute Learning Session.
Test #3: This test was conducted while standing.
Conclusion: Yarnkey has an error rate around 2% after 40 minutes of usage.
(seated)
SUS[10] Survey Results
18
Survey result #1: Learning Difficulty
Survey result #2: Ease-of-use
# of responses
# of responses
not at all definitely
not at all definitely
Questions from the SUS were used in the survey.
Conclusion: The results show that the system is easy to use as it has scores of 4+ for many questions.
Error Rate for Each Character
Conclusion: Analysis and development of a more ergonomic stroke pattern is needed
19
Overview of Error Rates for each Character�(higher rank is better)
v
p
r�f
Feedback from Users
One user mentioned potential uses:
One of the neat things about Yarnkey is its stealth—you can imagine a ... field scientist finding that feature useful. ... I can also imagine it being very popular as a game controller, for use on a bus or train.
20
Useful for quickly taking down digital notes
Conclusions and Future Work
Conclusions
Future Work
21
Acknowledgments
This work was supported by the JST Global Science Campus Experts in Information Science program hosted by the National Institute in Informatics, Information Processing Society of Japan, and the Japanese Committee for the International Olympiad in Informatics.
We appreciate the mentors and students in the program, Abby Kim and Kazuhiro Shinoda for their useful feedback.
Financial support was provided by the Masason Foundation.
22
References
[1]: TelemetRing: A Batteryless and Wireless Ring-shaped Keyboard using Passive Inductive Telemetry� doi:10.1145/3379337.3415873
[2]: TipText: Eyes-Free Text Entry on a Fingertip Keyboard doi:10.1145/3332165.3347865
[3]: A text input method for half-sized keyboard using keying interval doi:10.1145/2406367.2406375
[4]: A Text Input Interface using a Portable Clavier for Pianists doi:10.1109/ISWC.2007.4373787
[5]: 篠田和宏, 原田珠華, 佐野由季, 安斉周, 矢谷浩司, “導電糸を用いたウェアラブルデバイス向け手書� き文字入力デバイス,” in 電子情報通信学会総合大会, 2021.
[6]: CircuitPython. (n.d.). Retrieved April 24, 2022, from https://circuitpython.org/
[7]: "Svelte • Cybernetically enhanced web apps." Svelte, https://svelte.dev/.
[8]: I. S. MacKenzie and R. W. Soukoreff, "Phrase Sets for Evaluating Text Entry Techniques," in CHI ’03 � Extended Abstracts on Human Factors in Computing Systems, ser. CHI EA ’03. New York, NY, � USA:vAssociation for Computing Machinery, 2003, p. 754–755. [On-line]. Available: � https://doi.org/10.1145/765891.765971
[9]: L Hamstra-Bletz and A W Bl ̈ote. Development of hand-writing in primary school: a longitudinal � study. Perceptual and Motor Skills, 90(3), Jan 1990.
[10]: "System Usability Scale." [Online]. Available: � https://www.usability.gov/how-to-and-tools/methods/system-usability-scale.html
[11]: M. Weiser, “The computer for the 21st century,” ACMSIGMOBILE mobile computing and � communications review, vol. 3, no. 3, pp. 3–11, 1999.
[12]: "Jacquard by Google - Levi's®." Google, https://atap.google.com/jacquard/products/levi-trucker/.
23
Conclusions and Future Work
Conclusions
Future Work
24