Title Blue light alarm for a better morning : the effects of blue light ...

Title Blue light alarm for a better morning : the effects of blue light exposure on wakeningSub TitleAuthor 藤本, 隆寛(Fujimoto, Takahiro)

加藤, 朗( Katō, Akira)Publisher 慶應義塾大学大学院メディアデザイン研究科

Publication year 2021Jtitle

JaLC DOIAbstract

Notes 修士学位論文. 2021年度メディアデザイン学第891号Genre Thesis or DissertationURL https://koara.lib.keio.ac.jp/xoonips/modules/xoonips/detail.php?koara_id=KO40001001-0000202

1-0891

慶應義塾大学学術情報リポジトリ(KOARA)に掲載されているコンテンツの著作権は、それぞれの著作者、学会または出版社/発行者に帰属し、その権利は著作権法によって保護されています。引用にあたっては、著作権法を遵守してご利用ください。

The copyrights of content available on the KeiO Associated Repository of Academic resources (KOARA) belong to the respective authors, academic societies, orpublishers/issuers, and these rights are protected by the Japanese Copyright Act. When quoting the content, please follow the Japanese copyright act.

Powered by TCPDF (www.tcpdf.org)

http://www.tcpdf.org

Master’s Thesis

Academic Year 2021

Blue Light Alarm for a Better Morning

— The Effects of Blue Light Exposure on Wakening —

Keio University

Graduate School of Media Design

Takahiro Fujimoto

A Master’s Thesis

submitted to Keio University Graduate School of Media Design

in partial fulfillment of the requirements for the degree of

Master of Media Design

Takahiro Fujimoto

Master’s Thesis Advisory Committee:

Professor Akira Kato (Main Research Supervisor)

Professor Hideki Sunahara (Sub Research Supervisor)

Master’s Thesis Review Committee:

Professor Akira Kato (Chair)

Professor Hideki Sunahara (Co-Reviewer)

Professor Kai Kunze (Co-Reviewer)

Abstract of Master’s Thesis of Academic Year 2021

Blue Light Alarm for a Better Morning

— The Effects of Blue Light Exposure on Wakening —

Category: Science / Engineering

Summary

Today, people do not have to step outside to work, and can stay up until the morn-

ing due to the ubiquitous proliferation of information technology and electronic

lights. This trend has become more prominent since the outbreak of COVID-19.

People are getting less exposure to natural light during the day and more exposure

to electronic lights at nighttime. Blue light is a blue spectrum of light visible to the

human eye. It is also contained in sunlight and light from our electronic displays

and is mostly beneficial for people during the day. However, sleep researchers

claim that exposure to blue light after dark inhibits the secretion of melatonin,

a sleep-inducing hormone, which leads to the disruption of our circadian rhythm.

Although there are many research about how blue light disrupts the onset of sleep,

there are not so many studies on the offset of sleep; waking up.

This paper will examine how blue light will affect the quality of waking up. A

blue LED (470nm wavelength) is used to expose five participants with blue light

one hour before their designated wake-up time while measuring the participant’s

heart rate, noise of the room, brightness of the room and the movement in their

own bed. The experiment’s duration was on average ten nights per participant.

While the hour-long blue light exposure, all the participants showed an increase

in body movement and heart rate fluctuation, implying that the participants were

experiencing sleep stages 1, 2 and REM sleep. Furthermore, from this study, the

participants showed positive responses to the ”light-alarm”, claiming that their

quality of waking up was increased. Some interesting reactions were observed such

as the light exposure in the morning stopping them from waking up mid-sleep and

making them feel sleepier at night.

i

Abstract

Keywords:

sleep, blue light, circadian rhythm, melatonin, sensing, heart rate, alarm

Keio University Graduate School of Media Design

Takahiro Fujimoto

ii

Contents

Acknowledgements ix

1 Introduction 1

1.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

1.2. Purpose of this Research . . . . . . . . . . . . . . . . . . . . . . . 2

1.3. Thesis Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

2 Related Works 4

2.1. Circadian Rhythm . . . . . . . . . . . . . . . . . . . . . . . . . . 4

2.1.1 Effect of Alcohol on Heart Rate and Sleep Stages . . . . . 6

2.2. Blue Light and sleep . . . . . . . . . . . . . . . . . . . . . . . . . 6

2.3. Light and Getting to Sleep . . . . . . . . . . . . . . . . . . . . . . 8

2.3.1 The Effects of Home Lighting on Getting to Sleep . . . . . 8

2.4. Blue Light Stimulus for a Better Morning . . . . . . . . . . . . . . 8

3 Approach 10

3.1. Concept . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

3.1.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

3.2. Experiment Design . . . . . . . . . . . . . . . . . . . . . . . . . . 12

3.2.1 Experiment Flow . . . . . . . . . . . . . . . . . . . . . . . 12

3.2.2 Network Composition . . . . . . . . . . . . . . . . . . . . . 13

3.2.3 Lamp . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

3.2.4 Sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

4 Results 16

4.1. Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

4.1.1 Flashing Blue Light . . . . . . . . . . . . . . . . . . . . . . 16

4.1.2 Heart Rate as a Measure to Observe Sleep Cycle . . . . . . 16

iii

Contents

4.1.3 Waking up to Blue Light . . . . . . . . . . . . . . . . . . . 16

4.2. Final Experiment . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

4.2.1 Evaluation of the Final Experiment . . . . . . . . . . . . . 19

4.2.2 Overview of Analysis . . . . . . . . . . . . . . . . . . . . . 19

4.2.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

5 Discussion 42

5.1. Possible Applications . . . . . . . . . . . . . . . . . . . . . . . . . 46

5.2. Future Works . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

5.3. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

References 49

Appendices 54

A. Instruction Manual for Experiment Setup . . . . . . . . . . . . . . 54

B. Leeds Sleep Evaluation Questionnaire . . . . . . . . . . . . . . . . 57

iv

List of Figures

3.1 Experiment Overview Diagram

Icon credit:

Freepik: Applewatch/Gyroscope/Lamp/Laptop/Mic/Router/Smartphone

Eucalyp: CdS Cell . . . . . . . . . . . . . . . . . . . . . . . . . . 11

3.2 wavelength Characteristics of OSB5XNE3X1S. . . . . . . . . . . 14

4.1 Author’s heart rate obtained during sleep using Apple Watch.

On the left, no alcohol was consumed.

On the right, 8 standard drinks were ingested before going to bed. 17

4.2 Box plot of sleep cycles. . . . . . . . . . . . . . . . . . . . . . . . 17

4.3 The area shaded blue is when the subject is being exposed to the

blue light. The time of awakening is marked in red. . . . . . . . 18

4.4 Participant 1: The summary of the average of LSEQ survey (Q6

- Q10) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

4.5 Results of Participant 1(1/2). The area shaded blue is when the

subject is being exposed to the blue light. The time of awakening

is marked in red. . . . . . . . . . . . . . . . . . . . . . . . . . . . 24



is marked in red. . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

4.7 Body movement for participant 1. Large body movements are

marked in green. Light exposure is shaded in blue. . . . . . . . . 25


- Q10) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29



is marked in red. . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

v

List of Figures



is marked in red. . . . . . . . . . . . . . . . . . . . . . . . . . . . 30



is marked in red. . . . . . . . . . . . . . . . . . . . . . . . . . . . 31




- Q10) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32



is marked in red. . . . . . . . . . . . . . . . . . . . . . . . . . . . 33



is marked in red. . . . . . . . . . . . . . . . . . . . . . . . . . . . 34


- Q10) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35



is marked in red. . . . . . . . . . . . . . . . . . . . . . . . . . . . 36



is marked in red. . . . . . . . . . . . . . . . . . . . . . . . . . . . 37



is marked in red. . . . . . . . . . . . . . . . . . . . . . . . . . . . 38




- Q10) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39

vi

List of Figures



is marked in red. . . . . . . . . . . . . . . . . . . . . . . . . . . . 40



is marked in red. . . . . . . . . . . . . . . . . . . . . . . . . . . . 41



A.1 Instruction Manual (EN) . . . . . . . . . . . . . . . . . . . . . . 55

A.2 Instruction Manual (JP) . . . . . . . . . . . . . . . . . . . . . . 56

B.1 Leeds Sleep Evaluation Questionnaire (EN) . . . . . . . . . . . . 57

B.2 Leeds Sleep Evaluation Questionnaire (JP) . . . . . . . . . . . . 58

vii

List of Tables

2.1 Sleep stage classification results [1] . . . . . . . . . . . . . . . . . 6

4.1 Participant Information . . . . . . . . . . . . . . . . . . . . . . . 21

4.2 Results Summary of Participant 1 . . . . . . . . . . . . . . . . . 22





viii

Acknowledgements

First of all, I would like to thank my supervisor Akira Kato, subsupervisors Hideki

Sunahara and Kai Kunze. I would also like to thank Bon, who was always open

to any questions and helped me throughout my days at KMD.

I am very grateful for all of the participants who were kind enough to take

part in my 10 days long experiment. Futa, my drinking buddy, thanks. I’m sure

you will never read my thesis, and I am not planning to tell you that I have

written your name here. But you’ve been a big support throughout this study!

I’m also grateful for all of my buddies at KMD. My life at this school would

not have been the same and will not be the same after graduation without you

guys. Big hand of applause to my friends in the project room (when it used to

be open), even you, the guy whom I’ve never spoken with! Thank you, Ryo, from

the student administration, you handsome bloke. You make life so much easier

for EVERYONE in this school. Thank you, random guy from SDM, we’ve walk

past each other so many times, even in the times of lockdown. I don’t know your

name, we’ve never greeted each other, but I hope the feeling we have is mutual.

Finally, I would like to thank my parents who has always supported my aca-

demic endeavours throughout my life. Thank you for spending so much money on

my education, I will make sure to make use of each and every bit of experience I

have accumulated in my life and pass it on to the next generation.babababababababababababababababababab

·······················································- cut here -·······················································The names of people I should not forget (in alphabetical order):

Chikuma, Fuko, Futa, Kiku, Maki, Matsuken, Raful, Ragnar, Taku, Ya-

mamu

·······················································- cut here -·······················································

ix

Chapter 1

Introduction

1.1. Introduction

Waking up is a pain for many. The human body has an internal clock which tells

the human when to be sleepy and when to be awake, called the circadian clock.

The synchronisation of this system is largely dependent on sun light. With the

prevalence of electronic lights, the humankind is spending more time indoors, and

less time outdoors. A survey which reached out to 2,002 individuals in 2018 found

that children aged 3-12 are spending 35% less time outside [2]. Light stimulus

causes the part of the brain called pineal gland to stop secreting a hormone called

melatonin. It’s said that older generations have an easier time waking up due to

a decrease in production of melatonin [3]. The decrease in melatonin secretion is

also the reason for mid-sleep wakening. Melatonin is what keeps people sleeping.

One statistic shows that people are not getting enough sleep. A survey asking

1,031 people showed that in 1942, most people spent 7 hours sleeping, in 2013,

only 40% of the population get less than 6 hours of sleep [4, 5].

The blue light contained in natural light is critical to adjust the circadian

rhythm. Since the Corona virus pandemic, people in the world are spending

up to 33% more time inside their homes [6] Implying that more people are getting

deprived of exposure to natural light. This is a recipe for a crooked circadian

rhythm.

This work takes a look into the effect of blue light emitted by an LED on

human sleep hypothesising that using blue light prior to wakening will make it

easier to ”wake up on the right side of the bed”. With more people spending

less time being exposed to natural light, if exposing to blue LED light has the

potential to remediate the human circadian rhythm, these findings may lead to a

vast application of blue LEDs.

1

1. Introduction 1.2. Purpose of this Research

1.2. Purpose of this Research

The purpose of this research is to quantitatively observe the effects of blue light on

increasing the quality of wakening. This study will design a device to take data

such as heart rate, body movement, any irregular sound and room brightness

from the participant’s own bedrooms while exposing the subject to blue light

emitted from an LED at the designated time. On wakening, the participants will

be asked to fill out a sleep quality evaluation questionnaire named Leeds Sleep

Evaluation Questionnaire(LSEQ) to examine from a subjective perspective on the

experiment.

1.3. Thesis Overview

• Chapter1

The purpose of this chapter is to ease the reader into the topic while in-

troducing the hypothesis. The second purpose is to give a briefing on this

thesis.

• Chapter2

This chapter reviews the current understanding of the major elements men-

tioned in this study. By grasping the essences of the papers referenced in

this chapter, the viability of the hypothesis will become clear.

• Chapter3

Chapter 3 will explain the concept of this research and revisit the hypothesis.

It will go in depth explaining the experiment design, the devices which were

assembled and how they work together.

• Chapter4

Including the various preliminary experiments, chapter 4 will explore the

core parameters which need to be set for the final experiment. The way

in which the data is processed and evaluated will also be explained in this

chapter. A brief explanation of the tools that were used will be provided

as well. At the End of the chapter, the results which were collected will be

summarised and listed.

2

1. Introduction 1.3. Thesis Overview

• Chapter5

This chapter will summarise the findings from this study. Some considera-

tions such as experimental limitations will be mentioned. Furthermore, the

possible applications of the findings and future works can be found here.

3

Chapter 2

Related Works

2.1. Circadian Rhythm

The circadian rhythm is mainly controlled by the suprachiasmatic nuclei(SCN)

in the anterior hypothalamus. This part of the hypothalamus is responsible for

regulating the secretion of the sleep related hormone, melatonin produced in the

pineal gland. Stimulus given to the body which acts as a cue to adjust the

circadian rhythm are called Zeitgebers [7]. The most important Zeitgeber is known

to be light stimulus [8].

Sleep Stages

During a single night’s sleep, there are five different stages of sleep. These are

broadly classified as Rapid Eye Movement(REM) and Non-Rapid Eye Move-

ment(NREM) sleep which will be explained below largely referring to a paper [9]

published in 1989. From top to bottom, the following list is in the order of shallow

to deep sleep.

• REM Sleep

This is the only stage of sleep where people dream. There are four to six

occurrences of REM sleep in a single night’s sleep, they become longer each

episode. In total, REM sleep comprises 20% to 25% of the time sleeping.

• NREM Stage 1

In the onset and offset of sleep, the phase is called stage 1. Its duration

is around 1 to 7 minutes and during this stage, the person is easily woken

up by subtle noise or olfactory stimuli. The person is not so responsive to

4

2. Related Works 2.1. Circadian Rhythm

visual stimulus at this stage. Stage 1 sleep comprises 2% to 7% of the time

sleeping.

• NREM Stage 2

Following stage 1, the sleeping person enters stage 2 sleep. It continues

for 10 to 25 minutes. The stimulus which was successful in waking up the

sleeping subject at stage 1 does not lead to awakening at this phase. Stage

2 sleep comprises 45% to 55% of the time sleeping.

• NREM Stage 3 / NREM Stage 4

In this phase, the muscle in the body is very relaxed. In the first cycle, there

is only a few minutes of stage 3 before stage 4 sleep. Stage 3 and stage 4

sleep accounts for 15% to 25% of sleep(stage 3: 3% to8%, stage 4: 10% to

15% ) [10].

Throughout a night’s sleep, the sleeper experiences the cycle of REM sleep and

NREM. The first cycle takes around 70 to 100 minutes. While the cycle following

the first takes around 90 to 120 minutes. Body movements during sleep usually

implies that the sleep stage is changing to a sleep stage closer to consciousness. It

is said that body movements indicate a stage 1 or stage 2 sleep. During sleep stages

3 and 4, the muscles relax completely, and there are close to no body movements.

Furthermore, during REM sleep, a phenomena called sleep atonia occurs, where

the body experiences a temporary paralysis. The sleepers may hallucinate an

alien mutilation at this time, phases like this is called sleep paralysis.

Heart Rate Variability and Sleep Cycle

Heart Rate Variability(HRV) refers to the observation of one’s change in their

heart rate over time. It is possible to monitor one’s sleep stages through HRV.

The physiological mechanism of change in hear rate is due to the influence of the

two nervous systems handling the autonomic function; sympathetic and parasym-

pathetic nervous systems respectively abbreviated as SNS and PNS [11, 12]. The

heart rate decreases during sleep stages 1, 2, 3, 4 with smaller signs of high fre-

quency variability in the heart rate. On the other hand, whilst REM sleep, the

heart rate increased with higher variability [13]. Furthermore, a study in 2021

5

2. Related Works 2.2. Blue Light and sleep

showed that monitoring HRV to determine sleep stages is entirely a feasible op-

tion with no statistically significant difference (table 2.1) to the PSG [1].

Evaluation of HRV on Classifying Sleep Stages

Stages Specificity Sensitivity Accuracy

Wake .77 .95 .93

Light Sleep .69 .67 .69

Slow Wave Sleep .91 .72 .87

REM .92 .60 .84

Table 2.1 Sleep stage classification results [1]

2.1.1 Effect of Alcohol on Heart Rate and Sleep Stages

Alcohol holds activities in the PNS down. In a study which looked into the influ-

ence of alcohol during sleep, they found that a increase in the duration of REM

and stage 1 sleep was found along with the increase in heart rate [14]. There is an-

other finding which should be noted about the effect of alcohol. When the subject

is administered with alcohol every day, within three nights, the body becomes tol-

erant to the effects of alcohol. However, when alcohol was not administered to the

subject following the daily dose, or when the subject has completely metabolised

the ethanol mid-sleep, the REM sleep increased. This phenomenon is also known

as a REM ”rebound effect” [10].

2.2. Blue Light and sleep

Blue Light and Melatonin

Melatonin is a kind of a hormone which is dispatched from the pineal gland,

making one tired and sleepy. With age, the amount of melatonin secretion declines

[15]. This is said to be the reason why older generations wake up earlier. Wave

6

2. Related Works 2.2. Blue Light and sleep

lengths between 446–477nm has been shown to have the most effect on melatonin

regulation [16]. A study in 1988 tested the difference in melatonin suppression with

white light with a brightness of 3000, 1000, 500, 350, 200 lux. The results showed

a decrease in melatonin secretion by 71%, 67%, 44%, 38%, 16% respectively [17].

In a research conducted in 2011, experiment subjects were exposed with two

different kinds of light between 2:00-3:30 AM to see a difference in the suppression

of melatonin. One light source was a 4000K fluorescent white light. The other

source of light was a blue LED with a peak bandwidth of 469nm [18]. This

research revealed that a narrow band blue light is more effective in suppressing

the secretion of melatonin.

Blue Light and Getting to Sleep

An experiment in 2016 compared reading a story from an iPad and a book before

getting to sleep. The subject’s brain waves were measured using EEG. The re-

searchers found impact in the EEG when the subjects were reading from an iPad,

with an average of 30 minutes delay in sleep onset [19]. Another study in the

same year examined the efficacy of enhancing melatonin secretion, sleep quality

and sleepiness at night by wearing blue light blocking glasses. The study had 12

volunteers wear blue light blocking lenses 2 hours prior to bedtime while reading

from a ”self-luminous portable device”. The results showed that on both sleepi-

ness and melatonin secretion, the spectacle wearers showed significant differences

to the control group [20]. On the other hand, a study in 2017 concluded that

there are not enough concrete evidence to support the benefits of wearing blue

light blocking spectacles hours before going to bed for sleep quality [21]. Referring

to the study referenced in chapter 2.2, the blue light emitted from the electronic

displays may not have been a strong enough light source to affect one’s circadian

rhythm. Anyhow, there were no studies found on the intensity (lux) of the blue

light source if the wavelength characteristics showed dominance in wavelengths of

446-477nm which is the wave length which affects humans the most [16].

7

2. Related Works 2.3. Light and Getting to Sleep

2.3. Light and Getting to Sleep

In 2013, at the Sleep and Chrono-biology Laboratory in University of Colorado

Boulder, a team of scientists observed the shift in the participant’s circadian

rhythm after camping in the mountains for a week through measuring melatonin

in their saliva. They found that being exposed to only natural light drastically

changed the participant’s circadian rhythms and concluded that a decrease in

exposure to natural light during the day time and an increase in exposure to

electrical lighting during the day time has shifted the circadian clock [22]. It is

not definitely stated in this study whether the lack of natural light exposure or

the increase in exposure to electrical lighting after dark was the causation of this

phenomena. However, it seems plausible that it is possible to shift one’s circadian

cycle using electrical light. Another study in 1989 used a 3000 and 1000 lux

white light which was known to be effective in suppressing melatonin secretion

at that time. They found that exposing the subjects to light at 2100-2200 did

not suppress much melatonin, however, when they exposed the subjects at 0000-

0100 and 0400-0500, the melatonin concentration was less than half of what it

was before the light exposure [23]. Based on the findings of this study, the light

may well be effective in reducing melatonin concentration when exposed before

the subject wakes up.

2.3.1 The Effects of Home Lighting on Getting to Sleep

In recent years, there are many studies which investigate the influence of lighting

before one goes to sleep, especially blue light. Overall, these researches mostly

concludes that electronic lighting in homes have a negative effect on sleepiness at

night [24,25].

2.4. Blue Light Stimulus for a Better Morning

There are many related works on how blue light prevents people from getting

tired before going to bed. There are also works which shine light on how the

increase/ decrease in exposure to natural light affects the human circadian rhythm.

However, the effect of blue light exposure in the morning on the waking up routine

8

2. Related Works 2.4. Blue Light Stimulus for a Better Morning

has not been surveyed enough. This study hypothesises that by delimiting the

accumulation of melatonin during the last cycle of sleep, the sleepers can avoid

sleep inertia - the groggy feeling people get right after wakening -. Below are the

core suppositions which support this hypothesis.

• High melatonin blood concentration on wakening is the source of sleep in-

ertia.

• Blue light inhibits the secretion of melatonin.

• Blue light which are emitted from LEDs are sufficient to cause the effect

stated above granted that it has enough brightness.

9

Chapter 3

Approach

3.1. Concept

Blue light is known to reduce the secretion of melatonin in our body. This study

will use this effect aiming to alleviate the occurrence and effect of sleep iner-

tia. Sleep inertia, also known as sleep drunkenness, is a state which immedi-

ately follows wakening, with symptoms such as disorientation and declined per-

formance [26]. A study in 2013 concluded that the main cause of sleep inertia is

excessive discharge of melatonin in our body, a sleep-inducing hormone.

The concept of this study is to use blue light to reduce the secretion of mela-

tonin prior to wakening, aiming to alleviate the occurrence and effect of sleep

inertia. This study will evaluate the efficacy of this experiment through sensor

measurement and a standard questionnaire, which will be discussed later in this

chapter.

3.1.1 Overview

A diagram of five major items in this experiment is shown in figure3.1.

1. Laptop

The laptop will be the timekeeper. A program will be running which re-

ceives data from the sensors and saves it with a timestamp. It also sends

”on” / ”off” signals to the lamp controller when the designated time comes.

2. Lamp

This lamp is controlled by the laptop via WiFi. Positioned to expose the

light near the participants pillow, the lamp is turned on one hour prior to

the participant’s designated time of wakening.

10

3. Approach 3.1. Concept

Lamp

Router

Sensors

Laptop

ESPPRESSIF

ENBoot

CLK D0 D1 15 2 40 16 17 5 18 19 GND 22RX TX21 23 GND

3V3ENVPVN3432 35332526271412D3 13D2 GNDCMD5VESPPRESSIF

EN Boot

CLKD0 D1

152

40

1617

518

19GND

22RX

TX21

23GND3V3

ENVP

VN34

3235

3325

2627

1412

D313

D2GND

CMD5V

Apple Watch

IPhone

Ack

on / off

data

UDP port:

22222・

22223・

22224・

Bluetooth

Figure 3.1 Experiment Overview Diagram

Icon credit:

Freepik: Applewatch/Gyroscope/Lamp/Laptop/Mic/Router/Smartphone

Eucalyp: CdS Cell

11

3. Approach 3.2. Experiment Design

3. Sensors

The sensing device is equipped with ESP32, a micro-controller with the ca-

pability to communicate via Bluetooth and WiFi. It retrieves and processes

the sensor data, serialises it, and sends it to the laptop through the router. 9

bytes of data is sent each transmission, which contains the sequence number,

microphone value, gyro-sensor value and the CdS cell value. The sequence

number is used to detect major lapses in the network communication.

4. Router

In this experiment, the various devices, namely the laptop, sensor board and

the Lamp communicates throughWiFi. The router takes care of leasing local

IP addresses, and routing the data to designated recipients.

5. Apple Watch

The Apple Watch is used to take the participant’s heart rate while asleep.

It is connected to the participant’s IPhone and the exported data is sent to

be analysed after the experiment.

3.2. Experiment Design

3.2.1 Experiment Flow

The participants are handed the experiment kit to take home. All of the goods

needed to participate in the experiment is inside with a set of instructions to setup

the devices. A total of ten trials will be conducted per participant.

• Initial Setup

Firstly, the lamp is fixed above the pillow while the gyro sensor is placed

underneath. The microphone and the CdS cell are placed so that it is not

blocked by any objects. Secondly, the participants are asked to plug in the

router, and to switch on the laptop. The participant does not need to con-

figure anything related to the WiFi since all of the devices are preconfigured

to connect to the router. On the desktop of the laptop, there will be a

program which the test subject is asked to execute.

12


• Single Iteration of Experiment

The participant starts the program on the laptop, which then asks the par-

ticipant to designate the time of wakening the following morning. Two

seconds after the participant confirms the input time, the program starts

collecting the sensor data.

When the program starts running, the participant is free to go to bed after

making sure that they are wearing the Apple Watch.

Upon waking up, they are asked to fill out a sheet of questionnaire.

3.2.2 Network Composition

Three devices, laptop, lamp, and the sensor, are connected to the router. When

they are connected, the router gives them an arbitrary local IP address (192.168.111.X)

through the Dynamic Host Configuration Protocol(DHCP) Server. All the de-

vices send their datagrams to 192.168.111.255, a broadcast address which sends

the packet to every node in the local network. The transport protocol used in this

experiment is User Datagram Protocol(UDP). As shown in figure 3.1, three UDP

ports are used in this setup.

• UDP port 22222

The laptop sends it’s ”on/off” signals to 192.168.111.255:22222(IP address:port)

and the only device in this setup who listens to port 22222 is the micro-

controller embedded in the lamp.

• UDP port 22223

Since the experiment is designed using UDP, due to the protocol’s nature,

some signals get lost in transmission unlike its sibling, the Transmission

Control Protocol. Hence, the 22223 port is used by the Lamp to send an

acknowledgement(ACK) packet whenever it receives the ”on/off” signal.

• UDP port 22224

Port 22224 is used by the micro-controller which is embedded in the sensor

circuit board. It continuously streams the sensed data to this port, which

the laptop is listening to.

13


Figure 3.2 wavelength Characteristics of OSB5XNE3X1S.

3.2.3 Lamp

[htbp] The light is connected to ESP32. When the laptop sends the ”on” signal

to the micro-controller, the micro controller starts sending pulse width modula-

tion(PWM) signals. The signal runs through a smoothing circuit to prevent the

signal from oscillating. The smoothed signal triggers the MOS-FET, running the

current through the LED. When the laptop sends the ”off” signal, the micro-

controller stops sending the signals to the MOS-FET, which leads to the LEDs

turning off. The LED that is used in this experiment is OSB5XNE3X1S. Accord-

ing to the data sheet, a single LED uses three watts at the recommended configu-

ration. It’s brightness is 35 lumens, with a radiation angle of 120◦. Three of these

LEDs are mounted on one device, roughly exposing the subject’s eyes with the

brightness of 200 - 300 lux of light if placed within half a metre of the participant’s

head. This value, according to the study referred to in chapter 2.2, theoretically

induces 16% to 38% decrease in melatonin secretion [17]. Furthermore, while the

study mentioned used a white light which wavelength characteristics are evened

out, this study uses an LED which has a wavelength characteristic of 470 nm

(see figure 3.2) being the dominant wavelength which is known to have a stronger

effect on decreasing melatonin secretion [16].

14


3.2.4 Sensors

There are three sensors on the sensor board:

• Gyro-sensor L3GD20H

The gyro-sensor is placed on the bed to sense the participant’s movement

during sleep.

This Gyro-sensor communicates with the micro controller through the I2C

protocol. It sends the angular data of x, y, and z axis. The micro-controller

100 samples, summing up the absolute value of each axis. Then, it takes

the average of the samples and loads it to the serialised data stack.

• Microphone SPU0414HR5H-SB

The microphone is used to sense any ”out of the ordinary” sound in the

environment. This is important to rule out the effect of the surrounding

noise during the experiment.

The micro-controller reads the data from the sensor through analogue input.

Similar to the gyro-sensor, 100 samples are taken, and the maximum value

is loaded to the data stack.

• CdS Cell

CdS Cells are electronic components composed of Cadmium Sulphide and

is also known as a photo resistor. Its conductivity increases when the cell is

exposed to light. In this study, it is used to measure the relative brightness

of the room that the participant is sleeping in. It can also show the exact

timing of when the blue light lamp is triggered.

One other sensor is used in this experiment is the optical heart rate sensor inside

Apple Watch. This is the main source of information to predict the sleep cycle

that the participant is going through. According to several attempts of recording

data during sleep, the Apple Watch takes a single measurement of heart rate every

15 seconds to 10 minutes.

15

Chapter 4

Results

4.1. Experiments

4.1.1 Flashing Blue Light

In this first trial, the light was configured to flash twice in a second on the arrival

of the designated time. The flashing light seemed to cause abrupt wakening, which

defeated the purpose of this study of reducing the effect of sleep inertia. After

testing out several frequencies combined with fading in/out effects, the increase

in the rate of change in brightness sensed through the eyelids tended to make

the stimulus more noticeable even when asleep. Another finding is that while a

repeated on/off sequences is quite noticeable, a single occurrence of a light turning

on is not. This is considered and implemented in the final experiment.

4.1.2 Heart Rate as a Measure to Observe Sleep Cycle

The main measurement that this study will be focusing on will be the heart rate

collected using the Apple Watch. Congruent with the findings of one of the related

works [14], data showed (see figure 4.1 right) that alcohol consumption leads to

the increase in heart rate. Moreover, the gradual descent of the heart rate was

seen, which may indicate the liver breaking down the alcohol. The graph in figure

4.1 (left) also seems reassuring that the heart rate measured using Apple Watch

gives an adequate reference of the sleep cycle (figure 4.2).

4.1.3 Waking up to Blue Light

In the next experiment, the light emitting device was set to stay on for an hour

before the audio alarm went off. Values from a Gyro Sensor and a CdS cell sensor

16

4. Results 4.1. Experiments

Figure 4.1 Author’s heart rate obtained during sleep using Apple Watch.

On the left, no alcohol was consumed.

On the right, 8 standard drinks were ingested before going to bed.

60min 45min90min75min70min45min

Figure 4.2 Box plot of sleep cycles.

17

4. Results 4.2. Final Experiment

were also collected during this trial. The gyro sensor was placed under the pillow,

while the CdS cell was placed on the bed board near the blue light source. From

the second day, the light was configured to turn on in the middle of the night

where the subject was assumed to be deep asleep and unlikely to wake up from

the light being turned on. Though the heart rate did not change drastically after

the exposure of blue light in this experiment, the values in the had a tendency of

gradually rising after the onset of blue light. This experiment also showed that

placing the gyro-sensor and the CdS cell sensors at the said positions is effective

in finding the subject’s body movement during sleep and the brightness of the

subject’s room.

00:00 01:00 02:00 03:00 04:00 05:00 06:00 07:00

60

80

100

BPM

Heart Rate

00:00 01:00 02:00 03:00 04:00 05:00 06:00 07:000

10000

20000

mov

emen

t Gyro

00:00 01:00 02:00 03:00 04:00 05:00 06:00 07:000

1000

2000

brig

htne

ss

Light

May13_2338_data

02:00 03:00 04:00 05:00 06:00 07:00 08:00 09:00 10:00 11:00

60

80BP

M

Heart Rate

02:00 03:00 04:00 05:00 06:00 07:00 08:00 09:00 10:00 11:000

10000

20000

mov

emen

t Gyro

02:00 03:00 04:00 05:00 06:00 07:00 08:00 09:00 10:00 11:000

1000

2000

brig

htne

ss Light

May15_0108_data

04:00 05:00 06:00 07:00 08:00 09:00 10:0060

80

100

BPM

Heart Rate

04:00 05:00 06:00 07:00 08:00 09:00 10:000

50001000015000

mov

emen

t Gyro

04:00 05:00 06:00 07:00 08:00 09:00 10:000

1000

2000

brig

htne

ss Light

May16_0304_data

01:00 02:00 03:00 04:00 05:00 06:00 07:00

75100125150

BPM

Heart Rate

01:00 02:00 03:00 04:00 05:00 06:00 07:000

20000

40000

60000

mov

emen

t Gyro

01:00 02:00 03:00 04:00 05:00 06:00 07:000

1000

2000

brig

htne

ss Light

May17_0038_data

Figure 4.3 The area shaded blue is when the subject is being exposed to the blue

light. The time of awakening is marked in red.

4.2. Final Experiment

To scale this experiment, the system explained with figure 3.1 in chapter 3.1.1 was

devised. The flow of the experiment, explained in chapter 3.2.1 is shown to the

participants using an instruction manual of two-pages A.1 A.2. The instruction

18


lists the contents of the bag on the first page, and on the second page, there are

instructions on the initial setup, each night and day. Essentially, the instructions

ask the participants to plug everything in, position the sensors at certain places,

put on the heart rate monitoring watch before they go to bed and not to forget

to fill in the questionnaire when they wake up.

4.2.1 Evaluation of the Final Experiment

There are two parts in evaluating this experiment. Each morning following the

experiment, the participants fill in the Leeds sleep Evaluation Questionnaire. This

is used to distinguish which experiment attempts were subjectively better for

the participants compared to others. From the sensor data, the movement of

the participant while asleep, the noisiness of the environment and the relative

brightness of the room that the participant is sleeping in will be quantified. Along

with these values, this study will look into the participant’s heart rate and how it

reacts to the exposure of blue light during sleep.

Leeds Sleep Evaluation Questionnaire

The Leeds Sleep Evaluation Questionnaire is a questionnaire designed by Par-

rot and Hindmarch and contains ten questions which is answered by striking a

line through a 100mm scale. It was designed to survey the participant’s subjec-

tive evaluation of sleepiness [27]. Four factors of sleep is measured through the

questions: getting to sleep, quality of sleep, awakening from sleep and behaviour

following wakefulness. The copy of the actual questionnaire used during the ex-

periment can be found in the appendix. In this study, questions through six to

ten (awakening from sleep and behaviour following wakefulness.) will be used to

determine if the blue light had any positive effect on waking up.

4.2.2 Overview of Analysis

This analysis compares the variability in heart rate which is a good indicator

of sleep cycles [12] with the change of the values retrieved from the light sensor

near wakening. The users will set the activation time of the light at their desired

time for each morning. The purpose of this study is to find the influence of the

19


light stimulus and to see if the blue light has any contribution to a gentle waking

routine.

The study’s initial plan was to sort the waking routines into ”good” groups and

”bad” groups based on the data gathered through the LSEQ survey then compare

the sensor values to find a pattern. However, to utilise the incomplete samples,

the sensor data for each of the mornings while the exposure of the blue light was

examined to find a pattern.

4.2.3 Results

Summary

The experiment was carried out by 5 participants, with duration of around 10

days each. The participants were not told that they needed to participate for 10

consecutive days. The table which shows which data was available on which dates

can be found below (Tables 4.2, 4.3 and 4.4). Many problems were found in the

data collection procedure when the results came, which will be explained in the

following passages.

Participant 1

Out of the 11 attempts from this participant, there were 6 nights which, due to

network problems, the complete data was not retrieved (figure 4.5 top right /

bottom right). There were 4 nights worth of data which had all the data available

for this participant. Regarding the data for the LSEQ questionnaires and the

heart rate data, there were 6 nights which could be used to see the tendencies.

Overall, a gradual increase in the heart rate even within the one-hour exposure of

blue light could be observed (figures 4.5 and 4.6) with increased body movements

(figure 4.7). Moreover, after a brief interview with participant 1 after the 10-day

long experiment, we found that the participant did not feel that they were affected

by the light at all. However, the participant did comment that they saw the blue

light emitted from the device before they heard the alarm several times in the

questionnaire comment area. The correlation of the sleep duration and the LSEQ

wake-up quality average was 0.46.

20


Demographics of the participants

Participant# Gender Age Note

1 F 25 Suffers from mild insomnia.

2 M 25 Occasionally wakes up late.

3 F 63Suffers from insomnia. Occasionally takes sleep

inducing medication.

4 M 65 Often wakes up multiple times mid-sleep.

5 F 32Usually sleeps late. + Needs to wake up at very

early mornings 6 days a week.

Table 4.1 Participant Information

13.46

47.96

3.543.82

32.54

67.3

11.58

49.7

22.8 27.02

450540 540

350

810

600540

600 600540

0

100

200

300

400

500

600

700

800

0102030405060708090

100

7-Jun 9-Jun 11-Jun 13-Jun 15-Jun 17-Jun 19-Jun 21-Jun

Min

utes

Scor

e /1

00

Day

LSEQ Summary of Participant 1

Figure 4.4 Participant 1: The summary of the average of LSEQ survey (Q6 -

Q10)

21


Participant 1: Results of Final Experiment

Questionnaire Gyro data HR data CdS data Night of

O O O O JUN8

O X X X JUN9

O X X X JUN10

O X X X JUN11

O O O O JUN12

O X X X JUN13

O X O O JUN14

O X O O JUN15

O X O O JUN16

X X X X JUN17

O O O O JUN18

O X X X JUN19

Table 4.2 Results Summary of Participant 1

22




O O O X JUN8

O O O O JUN9

O O O O JUN10

X O O O JUN11

X O O O JUN12

O O O O JUN13

X X X X JUN14

X X X X JUN15

O O O X JUN16

O O O X JUN17

O O O X JUN18

X X X X JUN19

O O O X JUN20

X X X X JUN21

O O O X JUN22


23


03:00 04:00 05:00 06:00 07:00 08:00 09:0050

75

100

125

BPM

Heart Rate

03:00 04:00 05:00 06:00 07:00 08:00 09:000

20000

40000

mov

emen

t Gyro

03:00 04:00 05:00 06:00 07:00 08:00 09:000

1000

2000

brig

htne

ss Light

Jun09_0002_data

01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:00 09:0050

75

100

125

BPM

Heart Rate

01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:00 09:000

20000m

ovem

ent Gyro

01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:00 09:000

25

50

75

brig

htne

ss Light

Jun10_0009_data

00:00 01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:00 09:00 10:0050

75

100

125

BPM

Heart Rate

00:00 01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:00 09:00 10:000

1000

2000

3000

mov

emen

t Gyro

00:00 01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:00 09:00 10:000

500

1000

1500

brig

htne

ss Light

Jun12_2324_data

02:00 03:00 04:00 05:00 06:00 07:00 08:00 09:00 10:00 11:0050

75

100

125

BPM

Heart Rate

02:00 03:00 04:00 05:00 06:00 07:00 08:00 09:00 10:00 11:00

2.9

3.0

3.1

mov

emen

t Gyro

02:00 03:00 04:00 05:00 06:00 07:00 08:00 09:00 10:00 11:000

500

1000

brig

htne

ss Light

Jun14_2321_data

Figure 4.5 Results of Participant 1(1/2). The area shaded blue is when the

subject is being exposed to the blue light. The time of awakening is marked in

red.

24


02:00 03:00 04:00 05:00 06:00 07:00 08:00 09:0050

75

100

125BP

MHeart Rate

02:00 03:00 04:00 05:00 06:00 07:00 08:00 09:00

2.9

3.0

3.1

mov

emen

t Gyro

02:00 03:00 04:00 05:00 06:00 07:00 08:00 09:000

100

200

300

brig

htne

ss

Light

Jun16_0106_data

03:00 04:00 05:00 06:00 07:00 08:00 09:00 10:0050

75

100

125

BPM

Heart Rate

03:00 04:00 05:00 06:00 07:00 08:00 09:00 10:00

2.9

3.0

3.1

mov

emen

t Gyro

03:00 04:00 05:00 06:00 07:00 08:00 09:00 10:000

1000

2000

brig

htne

ss Light

Jun17_0154_data

03:00 04:00 05:00 06:00 07:00 08:00 09:00 10:0050

75

100

125

BPM

Heart Rate

03:00 04:00 05:00 06:00 07:00 08:00 09:00 10:000

10000

20000

mov

emen

t Gyro

03:00 04:00 05:00 06:00 07:00 08:00 09:00 10:000

1000

2000

brig

htne

ss Light

Jun19_0139_data

Jun20_0110_data



red.

02:00 03:00 04:00 05:00 06:00 07:00 08:00 09:00 10:00time

50

60

70

80

90

100

110

BPM

June 8th - 9th Heart Rate

Figure 4.7 Body movement for participant 1. Large body movements are marked

in green. Light exposure is shaded in blue.

25




X X X X JUN8

O X O O JUN9

O X X X JUN10

O X X X JUN11

O X X X JUN12

O X O O JUN13

O X O O JUN14

O X O O JUN15

O X O X JUN16

O X O X JUN17

O X O O JUN18


Participant 2

Out of the 15 days that participant 2 took part in this experiment, there were

4 days in which the participant could not initiate the device. Some days the

participant was not home, and other days the participant was too tired and/or

intoxicated to be able to carry out the experiment routine. Furthermore, for all

the remaining 11 days, the questionnaire, the data retrieved from the gyro sensor

and the heart rate data were available. The brightness data from the CdS cell

were not so perfect with only five days where the activation of the light emitting

device was observed. When the heart rate data and the time of when participant 2

was exposed to the blue light were compared, 8 mornings out of the 11 recordings

showed fluctuations in the heart rate while the participant was being exposed to

the light (figures 4.9, 4.10, 4.11). Large body movements were observed through

the sensor for most of the nights (figure 4.12). Overall, this participant claimed

that using the blue light device made their awakenings more pleasant compared

to standard morning wake up routines. Also, This particular participant did not

set an alarm for most of the mornings during this experiment, nevertheless, the

26




O O O O JUN22

O O O O JUN23

O O O O JUN24

O O O O JUN25

O O O O JUN26

O X X X JUN27

X X X X JUN28

O O O O JUN29

O O O O JUN30

O O O O JUL1

O X O O JUL2

O X O O JUL3

O X O O JUL4

O X O O JUL5


27




O O O O JUL5

O O O O JUL6

O O O O JUL7

X O O O JUL8

O O O O JUL9

O O O O JUL10

O O O O JUL11

X O X O JUL12

O O X O JUL13

O O X O JUL14

O O X O JUL15


participant was able to wake up most mornings. The participant stated that the

light emitting device woke them up without failure, although, they tended to go

back to sleep from the lack of abruptness and shock. The correlation of the sleep

duration and the LSEQ wake up quality average was -0.34.

Participant 3

The total duration of the participant 3’s experiment was 11 days. The light was

not configured to turn on and the data could not be retrieved for the first night

shown in table 4.4. Consequently, the participant was told to extend the date

for a day and since the light was not activated, the first morning’s questionnaire

was waived. This participant took part in the experiment for the following 10

consecutive days, and differently from participants 1 and 2, this participant set

the activation time of the light emitting device at similar times. After a few days

into the experiment, the participant delayed the activation time setting by 30

minutes since they noticed that they are woken up by the light after 30 minutes of

exposure. The questionnaire was filled for all attempts. However, all the values of

28


54.4

74.84

97.16

69.1658.98

91.28

26.26

89.56 90.56

36.7

360420

360

570

480

345 360300

360

480

300

400

500

600

700

800

900

0

20

40

60

80

100

120

7-Jun 9-Jun 11-Jun 13-Jun 15-Jun 17-Jun 19-Jun 21-Jun 23-Jun

Min

utes

Scor

e /1

00

Day



Q10)

02:00 03:00 04:00 05:00 06:00 07:00 08:0050

75

100

125

BPM

Heart Rate

02:00 03:00 04:00 05:00 06:00 07:00 08:000

20000

40000

mov

emen

t Gyro

02:00 03:00 04:00 05:00 06:00 07:00 08:000

50010001500

brig

htne

ss Light

Jun09_0107_data

01:00 02:00 03:00 04:00 05:00 06:00 07:0050

75

100

125

BPM

Heart Rate

01:00 02:00 03:00 04:00 05:00 06:00 07:000

2000

4000

6000

mov

emen

t Gyro

01:00 02:00 03:00 04:00 05:00 06:00 07:000

200

brig

htne

ss Light

Jun10_0014_data

01:00 02:00 03:00 04:00 05:00 06:0050

75

100

125

BPM

Heart Rate

01:00 02:00 03:00 04:00 05:00 06:000

20000

40000

mov

emen

t Gyro

01:00 02:00 03:00 04:00 05:00 06:000

200

400

600

brig

htne

ss Light

Jun11_0013_data

03:00 04:00 05:00 06:00 07:00 08:00 09:0050

75

100

125

BPM

Heart Rate

03:00 04:00 05:00 06:00 07:00 08:00 09:000

2000

4000

6000

mov

emen

t Gyro

03:00 04:00 05:00 06:00 07:00 08:00 09:000

250

500

750

brig

htne

ss Light

Jun12_0226_data



red.

29


02:00 04:00 06:00 08:00 10:00 12:0050

75

100

125

BPM

Heart Rate

02:00 04:00 06:00 08:00 10:00 12:000

10000

20000

30000

mov

emen

t Gyro

02:00 04:00 06:00 08:00 10:00 12:000

500

1000

1500

brig

htne

ss Light

Jun13_0042_data

01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:0050

75

100

125

BPM

Heart Rate

01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:000

20000

40000

mov

emen

t Gyro

01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:000

500

1000

1500

brig

htne

ss Light

Jun14_0003_data

01:00 02:00 03:00 04:00 05:0050

75

100

125

BPM

Heart Rate

01:00 02:00 03:00 04:00 05:00400

500

600

700

mov

emen

t Gyro

01:00 02:00 03:00 04:00 05:000

500

1000

brig

htne

ss

Light

Jun16_2358_data

00:00 01:00 02:00 03:00 04:00 05:00 06:0050

75

100

125

BPM

Heart Rate

00:00 01:00 02:00 03:00 04:00 05:00 06:00

400

500

600

mov

emen

t Gyro

00:00 01:00 02:00 03:00 04:00 05:00 06:000

100

200

brig

htne

ss Light

Jun17_2353_data



red.

30


05:00 06:00 07:00 08:00 09:0050

75

100

125

BPM

Heart Rate

05:00 06:00 07:00 08:00 09:00

400

500

600

mov

emen

t

Gyro

05:00 06:00 07:00 08:00 09:000.05

0.00

0.05

brig

htne

ss Light

Jun19_0420_data

02:00 03:00 04:00 05:00 06:00 07:00 08:0050

75

100

125

BPM

Heart Rate

02:00 03:00 04:00 05:00 06:00 07:00 08:000

20000

40000

60000

mov

emen

t Gyro

02:00 03:00 04:00 05:00 06:00 07:00 08:000

500

1000

brig

htne

ss Light

Jun21_0147_data

01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:0050

75

100

125

BPM

Heart Rate

01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:000

5000

10000

mov

emen

t Gyro

01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:000

1000

2000

brig

htne

ss Light

Jun23_0025_data



red.

00:00 01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:00time

60

65

70

75

80

85

BPM




31


65.96

54.2447.14

61.2

38.96

62

76.32

54.8

74.76

51.04

370330

266

423

310345 365

315

430 436

200250300350400450500550600650700

0102030405060708090

7-Jun 9-Jun 11-Jun 13-Jun 15-Jun 17-Jun 19-Jun 21-Jun

Min

utes

Scor

e /1

00

Day



Q10)

the gyro sensor continuously read a value of 3 (see figures 4.14 and 4.15), which is

an error code for the I2C interface. The CdS cell was not picking up light for three

consecutive days, and then two consecutive days three days apart and this was due

to an object covering the sensor up. The heart rate data from this participant

also showed a gradual increase while the exposure of the light. Participant 3

described the light to have the effect of starting the day with a bright mood.

Also, in the comments, participant 3 noted that since the experiment started,

they were waking up with a clear head and was able to wake up steadily despite

being deprived of sleep. They also said ”having something to contribute just by

sleeping gives me a reason to go to bed.”. The correlation of the sleep duration

and the LSEQ wake up quality average was 0.53.

Participant 4

The duration of participant 4’s experiment was 14 days. There was a night where

the sensor board had connection problems with the computer (table 4.5, JUL27).

The following day, the participant did not take any data since the participant

was not home. This participant set the activation time of the light at 5:30 AM

regardless of when the participant went to bed. The participant had sleeping

32


02:00 03:00 04:00 05:00 06:00 07:0050

75

100

125

BPM

Heart Rate

02:00 03:00 04:00 05:00 06:00 07:00

2.9

3.0

3.1

mov

emen

t Gyro

02:00 03:00 04:00 05:00 06:00 07:000

1000

2000

3000

brig

htne

ss Light

Jun10_0133_data

01:00 02:00 03:00 04:00 05:00 06:00 07:0050

75

100

125

BPM

Heart Rate

01:00 02:00 03:00 04:00 05:00 06:00 07:00

2.9

3.0

3.1

mov

emen

t Gyro

01:00 02:00 03:00 04:00 05:00 06:00 07:000

1000

2000

brig

htne

ss Light

Jun14_0018_data

03:00 04:00 05:00 06:00 07:0050

75

100

125

BPM

Heart Rate

03:00 04:00 05:00 06:00 07:00

2.9

3.0

3.1

mov

emen

t Gyro

03:00 04:00 05:00 06:00 07:000

1000

2000

brig

htne

ss Light

Jun15_0227_data

02:00 03:00 04:00 05:00 06:00 07:0050

75

100

125

BPM

Heart Rate

02:00 03:00 04:00 05:00 06:00 07:00

2.9

3.0

3.1

mov

emen

t Gyro

02:00 03:00 04:00 05:00 06:00 07:000

1000

2000

brig

htne

ss Light

Jun16_0154_data



red.

33


03:00 04:00 05:00 06:00 07:00 08:00 09:0050

75

100

125

BPM

Heart Rate

03:00 04:00 05:00 06:00 07:00 08:00 09:00

2.9

3.0

3.1

mov

emen

t Gyro

03:00 04:00 05:00 06:00 07:00 08:00 09:000.05

0.00

0.05

brig

htne

ss Light

Jun17_0235_data

03:00 04:00 05:00 06:00 07:00 08:0050

75

100

125

BPM

Heart Rate

03:00 04:00 05:00 06:00 07:00 08:00

2.9

3.0

3.1

mov

emen

t Gyro

03:00 04:00 05:00 06:00 07:00 08:000.05

0.00

0.05

brig

htne

ss Light

Jun18_0235_data

01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:0050

75

100

125

BPM

Heart Rate

01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:00

2.9

3.0

3.1

mov

emen

t Gyro

01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:000

1000

2000

3000

brig

htne

ss Light

Jun19_0029_data



red.

34


38.442.22

47.2851.29

58.7 59.14 59.9 58.76

47.3249.54

37.14410 419

335

422

324

250304

575

472

200250300350400450500550600650700

0

10

20

30

40

50

60

70

21-Jun 23-Jun 25-Jun 27-Jun 29-Jun 1-Jul 3-Jul 5-Jul

Min

utes

Scor

e /1

00

Day



Q10)

problems where they would wake up mid sleep. For this reason, there were con-

siderable days where the participant was awake before the light being activated.

To not stress the participant out about being asleep when the light was activated,

the participant was asked to try exposing themself to the light even if they were

awake. The gyro sensor seemed to have stopped working on the last four days of

the experiment. For the mornings when the participant was sleeping when the

light was activated, an increase in body movement (figure 4.20) and heart rate

fluctuation was observed (figures 4.17, 4.18 and 4.19). The correlation between

the sleep duration and the waking up quality for the LSEQ survey was negative

0.64. Participant 4 felt that the light had helped them sleep through the whole

night without waking up for some of the nights.

Participant 5

For participant 5, the duration of the experiment was 11 days. For two of the days,

the questionnaire could not be retrieved (table 4.6). Also, the Apple Watch which

is responsible for measuring heart rate data stopped logging values after the night

of JUL 11th. As the watch’s real-time heart rate monitor was working, the cause

is unknown. This participant usually woke up in early mornings, with an average

35


00:00 01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:00

40

60

80

BPM

Heart Rate

00:00 01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:000

5000

10000

mov

emen

t Gyro

00:00 01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:000

1000

2000

brig

htne

ss Light

Jun22_2356_data

01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:00

40

60

80

BPM

Heart Rate

01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:000

100002000030000

mov

emen

t Gyro

01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:000

1000

2000

brig

htne

ss Light

Jun24_0054_data

01:00 02:00 03:00 04:00 05:00 06:00 07:00

40

60

80

BPM

Heart Rate

01:00 02:00 03:00 04:00 05:00 06:00 07:000

20000

40000

60000

mov

emen

t Gyro

01:00 02:00 03:00 04:00 05:00 06:00 07:000

1000

2000

brig

htne

ss Light

Jun25_0019_data

00:00 01:00 02:00 03:00 04:00 05:00 06:00 07:00

40

60

80

BPM

Heart Rate

00:00 01:00 02:00 03:00 04:00 05:00 06:00 07:000

10000

20000

mov

emen

t Gyro

00:00 01:00 02:00 03:00 04:00 05:00 06:00 07:000

1000

2000

brig

htne

ss Light

Jun25_2355_data



red.

36


00:00 01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:00

40

60

80

BPM

Heart Rate

00:00 01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:000

5000

10000

mov

emen

t Gyro

00:00 01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:000

1000

2000

brig

htne

ss Light

Jun26_2333_data

02:00 03:00 04:00 05:00 06:00 07:00 08:00

40

60

80

BPM

Heart Rate

02:00 03:00 04:00 05:00 06:00 07:00 08:000

10000

20000

mov

emen

t Gyro

02:00 03:00 04:00 05:00 06:00 07:00 08:000

1000

2000

brig

htne

ss Light

Jun30_0101_data

01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:00 09:00

40

60

80

BPM

Heart Rate

01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:00 09:000

5000

10000

mov

emen

t Gyro

01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:00 09:000

1000

2000

brig

htne

ss Light

Jul01_0003_data

23:00 00:00 01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:00

40

60

80

BPM

Heart Rate

23:00 00:00 01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:000

10000

20000

mov

emen

t Gyro

23:00 00:00 01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:000

1000

2000

brig

htne

ss Light

Jul01_2253_data



red.

37


00:00 01:00 02:00 03:00 04:00 05:00 06:00 07:00

40

60

80BP

M

Heart Rate

00:00 01:00 02:00 03:00 04:00 05:00 06:00 07:000

10000

20000

30000

mov

emen

t Gyro

00:00 01:00 02:00 03:00 04:00 05:00 06:00 07:000

1000

2000

brig

htne

ss Light

Jul02_2332_data

01:00 02:00 03:00 04:00 05:00 06:00 07:00

40

60

80

BPM

Heart Rate

01:00 02:00 03:00 04:00 05:00 06:00 07:0033490

33500

33510

mov

emen

t Gyro

01:00 02:00 03:00 04:00 05:00 06:00 07:000

1000

2000

brig

htne

ss Light

Jul04_0004_data

00:00 01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:00

40

60

80

BPM

Heart Rate

00:00 01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:0032000

33000

34000

35000

mov

emen

t Gyro

00:00 01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:000

1000

2000

brig

htne

ss Light

Jul04_2305_data

01:00 02:00 03:00 04:00 05:00 06:00 07:00

40

60

80

BPM

Heart Rate

01:00 02:00 03:00 04:00 05:00 06:00 07:0032000

33000

34000

35000

mov

emen

t Gyro

01:00 02:00 03:00 04:00 05:00 06:00 07:000

1000

2000br

ight

ness Light

Jul06_0055_data



red.



38


97.32

82.9887.74

55.7868.16

97.6

83.3896.6 97.52

317365

417

285

450

366

290 306330

200250300350400450500550600650700

0

20

40

60

80

100

120

4-Jul 6-Jul 8-Jul 10-Jul 12-Jul 14-Jul 16-Jul

Min

utes

Scor

e /1

00

Day



Q10)

sleep duration of 5 hours and 47 minutes. The correlation of sleep duration and

wake up quality was negative 0.07 (figure 4.21). The participant claimed that they

usually require 6 hours of sleep when the participant can rest for two to three hours

before going to bed before midnight. During this experiment, the participant was

overloaded with work, causing them to sleep around 1AM- 2AM every day. This

may be the reason for the sleep duration having an extremely low correlation to

the wake up quality. This participant described the light to have had the effect

of making waking up in mornings easier. The data consistently showed increase

in body movement and an episode of increased heart rate fluctuations while the

exposure of blue light.

39


01:00 02:00 03:00 04:00 05:00 06:00

40

60

80

BPM

Heart Rate

01:00 02:00 03:00 04:00 05:00 06:000

10000

20000

mov

emen

t Gyro

01:00 02:00 03:00 04:00 05:00 06:000

500

1000

1500

brig

htne

ss

Light

maki_Jul06_0006_data

00:00 01:00 02:00 03:00 04:00 05:00 06:00

40

60

80

BPM

Heart Rate

00:00 01:00 02:00 03:00 04:00 05:00 06:000

50001000015000

mov

emen

t Gyro

00:00 01:00 02:00 03:00 04:00 05:00 06:000

1000

2000

brig

htne

ss Light


01:00 02:00 03:00 04:00 05:00 06:00 07:00

40

60

80

BPM

Heart Rate

01:00 02:00 03:00 04:00 05:00 06:00 07:000

10000

20000

30000

mov

emen

t Gyro

01:00 02:00 03:00 04:00 05:00 06:00 07:00

2000

2200

brig

htne

ss Light


02:00 03:00 04:00 05:00 06:00

40

60

80

BPM

Heart Rate

02:00 03:00 04:00 05:00 06:000

2500

5000

7500

mov

emen

t Gyro

02:00 03:00 04:00 05:00 06:000

50010001500

brig

htne

ss Light




red.

40


02:30 03:00 03:30 04:00 04:30 05:00 05:30 06:00 06:30

40

60

80BP

M

Heart Rate

02:30 03:00 03:30 04:00 04:30 05:00 05:30 06:00 06:300

5000

10000

mov

emen

t Gyro

02:30 03:00 03:30 04:00 04:30 05:00 05:30 06:00 06:300

1000

2000

3000

brig

htne

ss Light


01:00 02:00 03:00 04:00 05:00 06:00 07:00

40

60

80

BPM

Heart Rate

01:00 02:00 03:00 04:00 05:00 06:00 07:00

2000

4000

mov

emen

t Gyro

01:00 02:00 03:00 04:00 05:00 06:00 07:000

1000

2000

3000

brig

htne

ss Light


01:00 02:00 03:00 04:00 05:00 06:00 07:00

40

60

80

BPM

Heart Rate

01:00 02:00 03:00 04:00 05:00 06:00 07:000

10000

20000

30000

mov

emen

t Gyro

01:00 02:00 03:00 04:00 05:00 06:00 07:000

2000

4000

brig

htne

ss Light




red.

00:00 01:00 02:00 03:00 04:00 05:00 06:00time

50

60

70

80

90

100

110

BPM




41

Chapter 5

Discussion

Summary

The main contribution of this study is to design a system which could measure

the sleep cycle of the subject at their own homes. When measuring sleep in a lab,

the subjects sense stress simply from being in an unfamiliar environment. This

may cause discrepancies to the participant’s usual sleep pattern. The goodness of

this system is that the measurement of the sleeping data can be taken from the

participant’s familiar comfortable bed with easily affordable, accessible sensors.

The experiment showed that the body movement data and the heart rate data

together show an imperfect but okay indication of the participant’s sleep stage.

Overall, there are not enough complete data samples to definitively state that

being exposed to blue light prior to awakening contributes to a better waking

experience. However, the data taken from the heart rate monitor and the few body

movement data values show that the light was effective in waking the participants

up. All the participants showed consistent patterns of increased body movements

and heart rate fluctuations. The increase in body movement shows that the

sleeper is in stage 1 or 2 of sleep. Past studies have revealed that ”bad” waking

experiences are often caused from waking up in sleep stages 3, 4 or REM sleep.

Sleep stages 1 and 2 are a good sleep phase to wake up at. Although, as the

data showed, many participants experienced fluctuations in the heart rate during

the exposure, implying REM sleep. This is not a good phase to wake up. Data

showed that the participants seemed to be shifting back and forth between sleep

stages 1, 2, consciousness and REM sleep. Furthermore, most of the time, the

fluctuation in the heart rate, indicating REM sleep, ended within the first 20

minutes of exposure which may have contributed to the better waking experience

for the participants. Moreover, from the comment box beside the questionnaire

42

5. Discussion

(found in B.1, B.2), there were promising comments such as:

• ”I do not need the alarm clock to wake myself up when I have this light set

up.”

• ”The light wakes me up gently.”

• ”Waking up bears no stress when I am woken up by this light. I don’t feel like

I’ve been woken up, I feel like I’ve just opened my eyes from a long blink.”

• ”I woke up 30 minutes before I set my alarm, this does not usually happen.”

• ”Compared to other mornings when I am deprived of sleep, I feel less groggy

even though I definitely have not had enough sleep.”

• ”I usually stay in bed because my body refuses to budge after I gain con-

sciousness, but that is not the case right now.”

• ”My heart palpitates on wakening most mornings, I have not felt it since I

started taking part in the experiment (Day10).”

One of the participants is still using the light emitting device every morning. At

this moment, it has already been over a month since the start of the experiment.

Before, the participant was using the light each morning, they would experience

palpitation (a sudden rise in heart rate) once every three days on waking up.

Although, the participant claims to not have experienced the palpitation ever

since they started using the light device. The same participant also commented

that they felt less groggy with the light in sleep deprived mornings. Although

further interview revealed that there was a sleepiness rebound after few hours.

This phenomenon is also seen in another study [28].

There were also some notable comments which could be developed into an

interesting future work.

• ”The light wakes me up so gently that I often go straight back to sleep, I

want something to prevent this.”

• ”I feel like I get sleepier at night.”

43

5. Discussion

• ”After a few days of being exposed to the light at the same time each morning,

I stopped waking up mid-sleep. I don’t know if the light is helping, or if it’s

another reason but I slept through the whole night!!”

Delay Before Waking

Participant 1 on average took 62 minutes from the onset of the light alarm to

wake up.


wake up.

Participant 3’s wake-up time could not be estimated since the sensor regarding

the body movement was failing. Although, the participant self-reported that the

average wake-up time for them was around 30 minutes.

Information on participant 4’s delay before waking following the blue light ex-

posure could not be deemed relevant since the participant was awake for most

mornings before the light was activated.


wake up.

Looking at the age difference in participant 1, 2 versus participant 3, the time

difference is in line with the study explaining the decline of melatonin production

with age [15]. Still, this experiment does have a small sample size making this

merely a suggestion. Participant 5 may have been more sensitive to the light

stimulus since they had important plans every morning during the experiment.

Note that ’wake up’ here is used to describe the occurrence of the initial large

movement in the gyro sensor data.

LSEQ Wake Up Average

There were too many data missing for participant 1 to picture the difference in

the heart rate, body movement with the timing of the light exposure.

For participant 2, the LSEQ Wake up Average scores had a negative correlation

with the sleeping duration. When the comments were examined to investigate the

reason, the factor was alcohol. When the participant had one too many drinks,

they tended to sleep longer, waking up with a hangover which led to the low

LSEQ Wake Up scores. Other than alcohol, the timing of when the blue light was

44

5. Discussion

activated seemed to have a relation with the LSEQ scores. When the heart rate

is increasing or is at the peak when the light is activated, the scores tended to be

high. Whereas, when the heart rate has just crossed over the peak when the light

was activated, the heart rate tended to create a sine curve like shape within the

light activated area on the graph and the scores tended to be lower.

For participant 3, the gyro sensor completely failed to work, so there is no way

to tell with precision when the subject woke up. However, the same movement of

heart rate compared with the light activation was witnessed.

Variances Between Activation Times

Participants 2, 3 and 4 were mostly setting their light activation time at similar

or fixed times and sleeping at similar times. On the other hand, participant 1

had a highly varying sleeping schedule with bedtimes frequently changing plus or

minus 2 hours. This may be one of the reasons why participant 1 did not feel that

they were influenced by the blue light. Interestingly, even for participant 1 who

did not feel the influence of the light, body movements were prevalent (when the

sensor was taking data properly.), indicating a possible reaction to the light.

Problems with the Sensors Used

Out of all the data collected, when the gyro sensor was properly picking up the

movements of the subjects, the graphs showed when the participants woke up and

realised that the light was turned on. The heart rate increased synchronously

with the gyro sensor peaks on most occasions and if the measurement was taken

for an extended period, one would be able to distinguish the sleeping cycle and

stages to see the change in duration with the participant being exposed with the

light. However, this study did not take measurements of when the participants

were not being exposed to the light so it may be difficult even then to see a clear

difference in the sleeping stages. One possible experiment design is to periodically

turn the light on and off, perhaps with a cycle of 90 minutes, to see if the sleeping

cycles start to synchronise with the light. This way, there would not be a need

to compare two experiment conditions. If it is the case that blue light has the

capability to manipulate or direct how the sleeping stages shift from one another,

the light alarm may be devised as a scientifically proven waking device.

45

5. Discussion 5.1. Possible Applications

The sensors not being able to collect appropriate data may be due to the lack of

explanation or a check sheet which explains further in detail about how they should

position the sensors. The participants were not told specifically about the light

sensor, which may have led to them placing the light sensor at ineffective locations.

Also, the participants were not specifically told to sleep alone each night. There

were cases where the gyro sensor measurements were picking up movements of

another person on the bed which could be estimated from the inconsistencies in

the heart rate change. The gyro sensor sometimes had a problem with the I2C

communication protocol right after being disconnected from the Wi-Fi network

and reconnecting. Retrospectively, the design of the device should have foreseen

this event and incorporated an LED indicator to show that the I2C interface is

having an error. Some of the data were not retrieved because the computer cut

off from the Wi-Fi network abruptly. Since none of the computers did not try to

reconnect to the network after being disconnected, it may be a problem caused

by the Debian OS which were running in the computers used for this study. Also,

the manufacturer of Apple Watch was deploying an update for Apple Watch OS

during this experiment. Three recordings of one of the participant’s data could

not be retrieved since the synchronisation with the iPhone had stopped until the

Apple Watch OS was installed. What seemed to have happened is that the data

which overflowed from the Watch’s internal storage had been discarded. Ideally,

all the data collected on a single night should be uploaded to a server on a daily

basis and plotted to a graph, so that the examiner has the ability to check if there

are any faults with the measurements.

5.1. Possible Applications

As an Alarm Clock Device

Throughout this study, many of the participants who experienced using the blue

light alarm claimed that they have experienced an easier waking up routine. Some

even claimed to have felt that they did not need an alarm clock in the morning

to wake up at the desired time, given that they have had sufficient time to sleep.

Moreover, a repeated usage of this device for a longer duration of time may have

the effect of adjusting the circadian rhythm. However, in this study, the devices

46

5. Discussion 5.1. Possible Applications

were only used for 10 days for each participant so further studies on the use of the

device for a lengthened period of time is necessary to be assertive on this matter.

Sensing Alcohol Influence During Sleeping Hours

As shown in figure 4.1, alcohol significantly changes the HRV during sleep. While

a healthy night of sleep has somewhat of a ”standard” outline of the HR graph,

the influence of alcohol is easily visible. A closer look into this may accommodate

a new sleep tech device to objectively measure the user’s sleeping quality. A

paper published in Nature in 2021 proposes a contactless heartbeat monitor [29].

A combination with this monitoring device and the light alarm may provide a

better user experience as well.

Equipped and Built into the Bedroom

Furthermore, a futuristic house where the whole bedroom is equipped with a blue

LED where it lights up to smoothly wake up the user at the programmed time

can be one ultimate application.

Jet-lag Proof Airplanes

Currently, the cabin inside air crafts developed by Airbus is equipped with full

colour LEDs predominantly for artistic means [30]. Lighting up the aircraft with

blue-light aiming to manipulate the passenger’s melatonin secretion may lead to

a better after-flight experience with less symptoms of jetlag.

Syncing the Circadian Rhythm of Submarine Crews

Not much information was found about the use of light with different wavelengths.

However, in a submarine where natural light is absent, the use of blue light at the

right moments may significantly boost the quality of life for submarine crews.

47

5. Discussion 5.2. Future Works

5.2. Future Works

To achieve the full implementation of the applications in section 5.1 and to grasp

the full capability of the effects of blue light on sleep cycles, the following aspects

should be investigated:

• Blue light exposed at a fixed time each morning, may have the capability to

boost melatonin secretion after dark.

• Blue light exposed at a fixed time each morning, may have the capability

stop one from waking up mid-sleep.

• melatonin concentration decrease with LED light source with varying inten-

sities should be examined.

• The timing, amount of time and the light intensity or exposure patterns

needed to boost melatonin secretion at the night.

• The individual difference of when the blue light starts taking effect while

the subject is sleeping, perhaps observable from the same experiment with

a larger population.

• Further studies on the relationship between melatonin concentration and

HRV during sleep.

• Examining the contactless heart rate monitor if it gives a fair measurement

of the heart rate.

5.3. Conclusion

For the people who participated in the experiment, data consistently showed in-

crease in body movements and an episode of heart rate fluctuation, albeit the

sample size being small. This indicates that the sleepers were in lighter sleep

stages, where it is recommended for the sleeper to wake up to avoid sleep inertia.

Considering the overall positive comments which were received from the subjects,

it is plausible that exposing blue light an hour before waking gives the subjects a

better waking up experience.

48

References

[1] Liisa Kuula and Anu-Katriina Pesonen. Heart rate variability and first-

beat method for detecting sleep stages in healthy young adults: Feasibil-

ity study. JMIR Mhealth Uhealth, 9(2):e24704, Feb 2021. URL: http:

//mhealth.jmir.org/2021/2/e24704/, doi:10.2196/24704.

[2] BusinessWire. Survey finds today’s children are spending 35% less time

playing freely outside, Sep 2018. URL: https://www.businesswire.

com/news/home/20180920005526/en/Survey-Finds-Today%E2%80%99s-

Children-Are-Spending-35-Less-Time-Playing-Freely-Outside.

[3] Ian Hickie, Joanne Carpenter, and Rebecca Robillard. Variations in the

sleep–wake cycle from childhood to adulthood: chronobiological perspec-

tives. ChronoPhysiology and Therapy, 2015:37, 07 2015. doi:10.2147/CPT.

S41765.

[4] Jeffrey M. Jones. In u.s., 40% get less than recommended amount of

sleep, Jun 2021. URL: https://news.gallup.com/poll/166553/less-

recommended-amount-sleep.aspx.

[5] Shawn Youngstedt, Eric Goff, Alexandria Reynolds, Daniel Kripke, Michael

Irwin, Richard Bootzin, Nidha Khan, and Girardin Jean-Louis. Has adult

sleep duration declined over the last 50+ years? Sleep Medicine Reviews, 28,

08 2015. doi:10.1016/j.smrv.2015.08.004.

[6] Research Ritchie and data: Hannah. Covid-19: Google mobility trends.

URL: https://ourworldindata.org/covid-google-mobility-trends.

[7] Wytske A. Hofstra and Al W. de Weerd. How to assess circadian rhythm

in humans: A review of literature. Epilepsy & Behavior, 13(3):438–

444, 2008. URL: https://www.sciencedirect.com/science/article/

49

http://mhealth.jmir.org/2021/2/e24704/

http://mhealth.jmir.org/2021/2/e24704/

http://dx.doi.org/10.2196/24704

https://www.businesswire.com/news/home/20180920005526/en/Survey-Finds-Today%E2%80%99s-Children-Are-Spending-35-Less-Time-Playing-Freely-Outside



http://dx.doi.org/10.2147/CPT.S41765

http://dx.doi.org/10.2147/CPT.S41765

https://news.gallup.com/poll/166553/less-recommended-amount-sleep.aspx

https://news.gallup.com/poll/166553/less-recommended-amount-sleep.aspx

http://dx.doi.org/10.1016/j.smrv.2015.08.004

https://ourworldindata.org/covid-google-mobility-trends

https://www.sciencedirect.com/science/article/pii/S1525505008001625

References

pii/S1525505008001625, doi:https://doi.org/10.1016/j.yebeh.2008.

06.002.

[8] Jeanne F. Duffy and Jr. Kenneth P. Wright. Entrainment of the human cir-

cadian system by light. Journal of Biological Rhythms, 20(4):326–338, 2005.

PMID: 16077152. URL: https://doi.org/10.1177/0748730405277983,

arXiv:https://doi.org/10.1177/0748730405277983, doi:10.1177/

0748730405277983.

[9] Mary Carskadon and William Dement. Normal human sleep: An overview.

principles and practice of sleep medicine. m.h. kryger (ed.). W.B. Saunders,

Philadelphia, pages 3–13, 01 1989.

[10] T. Lee-Chiong. Sleep: A Comprehensive Handbook. 11 2005. doi:10.1002/

0471751723.

[11] Phyllis K. Stein and Yachuan Pu. Heart rate variability, sleep and sleep

disorders. Sleep Medicine Reviews, 16(1):47–66, 2012. URL: https:

//www.sciencedirect.com/science/article/pii/S1087079211000293,

doi:https://doi.org/10.1016/j.smrv.2011.02.005.

[12] Armin Bunde, Shlomo Havlin, Jan W. Kantelhardt, Thomas Penzel, Jorg-

Hermann Peter, and Karlheinz Voigt. Correlated and uncorrelated regions

in heart-rate fluctuations during sleep. Phys. Rev. Lett., 85:3736–3739, Oct

2000. URL: https://link.aps.org/doi/10.1103/PhysRevLett.85.3736,

doi:10.1103/PhysRevLett.85.3736.

[13] Danguole ZEmaityte, Giedrius Varoneckas, and Eugene Sokolov. Heart

rhythm control during sleep. Psychophysiology, 21(3):279–289, 1984.

URL: https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1469-

8986.1984.tb02935.x, arXiv:https://onlinelibrary.wiley.

com/doi/pdf/10.1111/j.1469-8986.1984.tb02935.x, doi:https:

//doi.org/10.1111/j.1469-8986.1984.tb02935.x.

[14] Yohei Sagawa, Hideaki Kondo, Namiko Matsubuchi, Takaubu Take-

mura, Hironobu Kanayama, Yoshihiko Kaneko, Takashi Kanbayashi,

Yasuo Hishikawa, and Tetsuo Shimizu. Alcohol has a dose-related effect on

50



http://dx.doi.org/https://doi.org/10.1016/j.yebeh.2008.06.002

http://dx.doi.org/https://doi.org/10.1016/j.yebeh.2008.06.002

https://doi.org/10.1177/0748730405277983

http://arxiv.org/abs/https://doi.org/10.1177/0748730405277983

http://dx.doi.org/10.1177/0748730405277983

http://dx.doi.org/10.1177/0748730405277983

http://dx.doi.org/10.1002/0471751723

http://dx.doi.org/10.1002/0471751723



http://dx.doi.org/https://doi.org/10.1016/j.smrv.2011.02.005

https://link.aps.org/doi/10.1103/PhysRevLett.85.3736

http://dx.doi.org/10.1103/PhysRevLett.85.3736

https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1469-8986.1984.tb02935.x

https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1469-8986.1984.tb02935.x

http://arxiv.org/abs/https://onlinelibrary.wiley.com/doi/pdf/10.1111/j.1469-8986.1984.tb02935.x

http://arxiv.org/abs/https://onlinelibrary.wiley.com/doi/pdf/10.1111/j.1469-8986.1984.tb02935.x

http://dx.doi.org/https://doi.org/10.1111/j.1469-8986.1984.tb02935.x

http://dx.doi.org/https://doi.org/10.1111/j.1469-8986.1984.tb02935.x

References

parasympathetic nerve activity during sleep. Alcoholism: Clinical and Exper-

imental Research, 35(11):2093–2100, 2011. URL: https://onlinelibrary.

wiley.com/doi/abs/10.1111/j.1530-0277.2011.01558.x, arXiv:https:

//onlinelibrary.wiley.com/doi/pdf/10.1111/j.1530-0277.2011.

01558.x, doi:https://doi.org/10.1111/j.1530-0277.2011.01558.x.

[15] Richard J. Wurtman. Age-Related Decreases in Melatonin Secretion―

Clinical Consequences. The Journal of Clinical Endocrinology & Metabolism,

85(6):2135–2136, 06 2000. URL: https://doi.org/10.1210/jcem.85.6.

6660, arXiv:https://academic.oup.com/jcem/article-pdf/85/6/2135/

9174440/jcem2135.pdf, doi:10.1210/jcem.85.6.6660.

[16] George C. Brainard, John P. Hanifin, Jeffrey M. Greeson, Brenda

Byrne, Gena Glickman, Edward Gerner, and Mark D. Rollag. Ac-

tion spectrum for melatonin regulation in humans: Evidence for a

novel circadian photoreceptor. Journal of Neuroscience, 21(16):6405–

6412, 2001. URL: https://www.jneurosci.org/content/21/16/6405,

arXiv:https://www.jneurosci.org/content/21/16/6405.full.pdf,

doi:10.1523/JNEUROSCI.21-16-06405.2001.

[17] Iain M. Mclntyre, Trevor R. Norman, Graham D. Burrows, and

Stuart M. Armstrong. Human melatonin suppression by light

is intensity dependent. Journal of Pineal Research, 6(2):149–156,

1989. URL: https://onlinelibrary.wiley.com/doi/abs/10.1111/j.

1600-079X.1989.tb00412.x, arXiv:https://onlinelibrary.wiley.com/

doi/pdf/10.1111/j.1600-079X.1989.tb00412.x, doi:https://doi.org/

10.1111/j.1600-079X.1989.tb00412.x.

[18] Kathleen E. West, Michael R. Jablonski, Benjamin Warfield, Kate S. Ce-

cil, Mary James, Melissa A. Ayers, James Maida, Charles Bowen, David H.

Sliney, Mark D. Rollag, John P. Hanifin, and George C. Brainard. Blue

light from light-emitting diodes elicits a dose-dependent suppression of mela-

tonin in humans. Journal of Applied Physiology, 110(3):619–626, 2011.

PMID: 21164152. URL: https://doi.org/10.1152/japplphysiol.01413.

2009, arXiv:https://doi.org/10.1152/japplphysiol.01413.2009, doi:

51

https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1530-0277.2011.01558.x

https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1530-0277.2011.01558.x

http://arxiv.org/abs/https://onlinelibrary.wiley.com/doi/pdf/10.1111/j.1530-0277.2011.01558.x



http://dx.doi.org/https://doi.org/10.1111/j.1530-0277.2011.01558.x

https://doi.org/10.1210/jcem.85.6.6660

https://doi.org/10.1210/jcem.85.6.6660

http://arxiv.org/abs/https://academic.oup.com/jcem/article-pdf/85/6/2135/9174440/jcem2135.pdf

http://arxiv.org/abs/https://academic.oup.com/jcem/article-pdf/85/6/2135/9174440/jcem2135.pdf

http://dx.doi.org/10.1210/jcem.85.6.6660

https://www.jneurosci.org/content/21/16/6405

http://arxiv.org/abs/https://www.jneurosci.org/content/21/16/6405.full.pdf

http://dx.doi.org/10.1523/JNEUROSCI.21-16-06405.2001

https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1600-079X.1989.tb00412.x

https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1600-079X.1989.tb00412.x

http://arxiv.org/abs/https://onlinelibrary.wiley.com/doi/pdf/10.1111/j.1600-079X.1989.tb00412.x

http://arxiv.org/abs/https://onlinelibrary.wiley.com/doi/pdf/10.1111/j.1600-079X.1989.tb00412.x

http://dx.doi.org/https://doi.org/10.1111/j.1600-079X.1989.tb00412.x

http://dx.doi.org/https://doi.org/10.1111/j.1600-079X.1989.tb00412.x

https://doi.org/10.1152/japplphysiol.01413.2009

https://doi.org/10.1152/japplphysiol.01413.2009

http://arxiv.org/abs/https://doi.org/10.1152/japplphysiol.01413.2009

http://dx.doi.org/10.1152/japplphysiol.01413.2009


References

10.1152/japplphysiol.01413.2009.

[19] Janne Grønli, Ida Kristiansen Byrkjedal, Bjørn Bjorvatn, Øystein Nødtvedt,

Børge Hamre, and Stale Pallesen. Reading from an ipad or from a book

in bed: the impact on human sleep. a randomized controlled crossover

trial. Sleep Medicine, 21:86–92, 2016. URL: https://www.sciencedirect.

com/science/article/pii/S1389945716000599, doi:https://doi.org/

10.1016/j.sleep.2016.02.006.

[20] Masahiko Ayaki, Atsuhiko Hattori, Yusuke Maruyama, Masaki Nakano, Mi-

chitaka Yoshimura, Momoko Kitazawa, Kazuno Negishi, and Kazuo Tsubota.

Protective effect of blue-light shield eyewear for adults against light pollution

from self-luminous devices used at night. Annual Review of Chronopharma-

cology, 33(1):134–139, January 2016. Publisher Copyright: © 2016 Taylor

& Francis. doi:10.3109/07420528.2015.1119158.

[21] John G Lawrenson, Christopher C Hull, and Laura E Downie. The ef-

fect of blue-light blocking spectacle lenses on visual performance, macu-

lar health and the sleep-wake cycle: a systematic review of the literature.

Ophthalmic and Physiological Optics, 37(6):644–654, 2017. URL: https://

onlinelibrary.wiley.com/doi/abs/10.1111/opo.12406, arXiv:https:

//onlinelibrary.wiley.com/doi/pdf/10.1111/opo.12406, doi:https:

//doi.org/10.1111/opo.12406.

[22] Kenneth P. Wright, Andrew W. McHill, Brian R. Birks, Brandon R.

Griffin, Thomas Rusterholz, and Evan D. Chinoy. Entrainment of

the human circadian clock to the natural light-dark cycle. Current

Biology, 23(16):1554–1558, 2013. URL: https://www.sciencedirect.

com/science/article/pii/S0960982213007641, doi:https://doi.org/

10.1016/j.cub.2013.06.039.

[23] Ian M. McIntyre, Trevor R. Norman, Graham D. Burrows, and Stuart M.

Armstrong. Human melatonin response to light at different times of the

night. Psychoneuroendocrinology, 14(3):187–193, 1989. URL: https://

www.sciencedirect.com/science/article/pii/0306453089900164, doi:

https://doi.org/10.1016/0306-4530(89)90016-4.

52





http://dx.doi.org/https://doi.org/10.1016/j.sleep.2016.02.006

http://dx.doi.org/https://doi.org/10.1016/j.sleep.2016.02.006

http://dx.doi.org/10.3109/07420528.2015.1119158

https://onlinelibrary.wiley.com/doi/abs/10.1111/opo.12406

https://onlinelibrary.wiley.com/doi/abs/10.1111/opo.12406

http://arxiv.org/abs/https://onlinelibrary.wiley.com/doi/pdf/10.1111/opo.12406

http://arxiv.org/abs/https://onlinelibrary.wiley.com/doi/pdf/10.1111/opo.12406

http://dx.doi.org/https://doi.org/10.1111/opo.12406

http://dx.doi.org/https://doi.org/10.1111/opo.12406



http://dx.doi.org/https://doi.org/10.1016/j.cub.2013.06.039

http://dx.doi.org/https://doi.org/10.1016/j.cub.2013.06.039

https://www.sciencedirect.com/science/article/pii/0306453089900164

https://www.sciencedirect.com/science/article/pii/0306453089900164

http://dx.doi.org/https://doi.org/10.1016/0306-4530(89)90016-4

http://dx.doi.org/https://doi.org/10.1016/0306-4530(89)90016-4

References

[24] Asuka Ishihara, Insung Park, Yoko Suzuki, Katsuhiko Yajima, Huiyun Cui,

Masashi Yanagisawa, Takeshi Sano, Junji Kido, and Kumpei Tokuyama.

Metabolic responses to polychromatic led and oled light at night. Scientific

reports, 11(1):1–11, 2021.

[25] Sean W Cain, Elise M McGlashan, Parisa Vidafar, Jona Mustafovska, Si-

mon PN Curran, Xirun Wang, Anas Mohamed, Vineetha Kalavally, and An-

drew JK Phillips. Evening home lighting adversely impacts the circadian

system and sleep. Scientific reports, 10(1):1–10, 2020.

[26] Patricia Tassi and Alain Muzet. Sleep inertia. Sleep Medicine Reviews,

4(4):341–353, 2000. URL: https://www.sciencedirect.com/science/

article/pii/S1087079200900984, doi:https://doi.org/10.1053/smrv.

2000.0098.

[27] J.A. Rosas and W.M. Anderson. Questionnaires and rating scales.

In Clete A. Kushida, editor, Encyclopedia of Sleep, pages 8–13. Aca-

demic Press, Waltham, 2013. URL: https://www.sciencedirect.

com/science/article/pii/B9780123786104001297, doi:https:

//doi.org/10.1016/B978-0-12-378610-4.00129-7.

[28] Henri Comtet, Pierre A Geoffroy, Mio Kobayashi Frisk, Jeffrey Hubbard, Lu-

divine Robin-Choteau, Laurent Calvel, Laurence Hugueny, Antoine U Viola,

and Patrice Bourgin. Light therapy with boxes or glasses to counteract effects

of acute sleep deprivation. Scientific reports, 9(1):1–9, 2019.

[29] Anran Wang, Dan Nguyen, Arun R Sridhar, and Shyamnath Gollakota. Us-

ing smart speakers to contactlessly monitor heart rhythms. Communications

biology, 4(1):1–12, 2021.

[30] AIRBUS. Led lighting (spectrum), 2021. URL: http://https://services.

airbus.com/en/in-flight-experience/cabin-upgrades/lighting-

comfort/led-lighting-spectrum.html.

53



http://dx.doi.org/https://doi.org/10.1053/smrv.2000.0098

http://dx.doi.org/https://doi.org/10.1053/smrv.2000.0098

https://www.sciencedirect.com/science/article/pii/B9780123786104001297

https://www.sciencedirect.com/science/article/pii/B9780123786104001297

http://dx.doi.org/https://doi.org/10.1016/B978-0-12-378610-4.00129-7

http://dx.doi.org/https://doi.org/10.1016/B978-0-12-378610-4.00129-7

http://https://services.airbus.com/en/in-flight-experience/cabin-upgrades/lighting-comfort/led-lighting-spectrum.html



Appendices

A. Instruction Manual for Experiment Setup

54

Appendices A. Instruction Manual for Experiment Setup

Figure A.1 Instruction Manual (EN)

55

Appendices A. Instruction Manual for Experiment Setup

Figure A.2 Instruction Manual (JP)

56

Appendices B. Leeds Sleep Evaluation Questionnaire

B. Leeds Sleep Evaluation Questionnaire

Lee

ds

Sle

ep E

valu

ati

on Q

ues

tion

na

ire

How

wou

ld y

ou d

escr

ibe

the

wa

y yo

u c

urr

entl

y fa

ll a

slee

p i

n c

omp

ari

son

to

usu

al?

-G

etti

ng

to

slee

p

How

wou

ld y

ou d

escr

ibe

the

qu

ali

ty o

f yo

ur

slee

p c

omp

are

d t

o n

orm

al

slee

p?

-Q

ua

lity

of

slee

p

How

wou

ld y

ou d

escr

ibe

you

r a

wa

ken

ing

in

com

pa

riso

n t

o u

sua

l? -

Aw

ake

fol

low

ing

sle

ep

How

do

you

fee

l w

hen

you

wa

ke u

p?

-A

wa

ke f

ollo

win

g s

leep

How

do

you

fee

l n

ow?

-B

eha

viou

r fo

llow

ing

wa

ken

ing

How

wou

ld y

ou d

escr

ibe

you

r b

ala

nce

an

d c

o-or

din

ati

on u

pon

aw

ake

nin

g?

-B

eha

viou

r fo

llow

ing

wa

ken

ing

1.

M

ore

dif

ficu

lt

tha

n u

sua

l

Ea

sier

th

an

usu

al

2.

S

low

er t

ha

n

usu

al

Mor

e q

uic

kly

tha

n u

sua

l

4.

M

ore

rest

less

tha

n u

sua

l

Ca

lmer

th

an

usu

al

5.

W

ith

mor

e

wa

kefu

l p

erio

ds

tha

n u

sua

l

Wit

h l

ess

wa

kefu

l p

erio

ds

tha

n u

sua

l

3.

I fe

el l

ess

slee

py

tha

n u

sua

l

Mor

e sl

eep

y

tha

n u

sua

l

6.

M

ore

dif

ficu

lt

tha

n u

sua

l

Ea

sier

th

an

usu

al

10

.

Mor

e d

isru

pte

d

th

an

usu

al

Les

s d

isru

pte

d

tha

n u

sua

l

9.

T

ired

Ale

rt

7.

R

equ

ires

a

per

iod

of

tim

e

lon

ger

th

an

usu

al

Sh

orte

r th

an

usu

al

8.

T

ired

Ale

rt

Wh

at

tim

e d

id y

ou *

go

to b

ed l

ast

nig

ht?

_

___:

____

Wh

at

tim

e d

id y

ou *

*get

up

fro

m b

ed t

oda

y? _

___:

____

* th

e ti

me

you

wen

t to

bed

, n

ot t

he

tim

e yo

u s

lep

t.

** t

he

tim

e yo

u g

ot u

p f

rom

bed

, n

ot w

hen

you

r

ega

ined

con

scio

usn

ess.

ID#

Da

y#

Fee

db

ack

on

th

e ex

per

ien

ce:

Figure B.1 Leeds Sleep Evaluation Questionnaire (EN)

57

Appendices B. Leeds Sleep Evaluation Questionnaire

Lee

ds

Sle

ep E

valu

ati

on Q

ues

tion

na

ire

寝つ

きに

つい

て

睡眠

の質

につ

いて

起床

時

起床

時の

気分

現在

の気

分

起床

時の

心身

の感

覚

1.

困

難容

易

2.

寝

付く

まで

に

時間

がか

かっ

た

特に

はや

く

寝つ

けた

4.

寝

苦し

かっ

た寝

やす

かっ

た

5.

い

つも

より

夜中

起き

るこ

とが

多か

った

いつ

もよ

り夜

中

起き

るこ

とが

少な

かっ

た

3.

ま

った

く眠

く

なか

った

とて

も

眠か

った

6.

困

難容

易

10

. 心

身の

バ

ラン

スが

悪い

心身

のバ

ラン

ス

がと

れて

いる

9.

疲

れて

いる

すっ

きり

して

いる

7.

起

床す

るま

でに

時間

がか

かっ

た

起床

する

まで

容易

8.

疲

れて

いる

すっ

きり

して

いる

昨晩

の*

就寝

時間

は何

時で

すか

?　

　　

:___

_

今朝

の**

起床

時間

は何

時で

すか

?

　　

:___

_

* 眠

りに

つい

た時

間で

はな

く、

ベッ

ドや

布団

に入

った

時間

。

** 目

が覚

めた

時間

では

なく

、ベ

ッド

や布

団か

ら起

き上

がっ

た時

間。

ID#

Da

y#

今回

の体

験に

つい

ての

コメ

ント

欄:

Figure B.2 Leeds Sleep Evaluation Questionnaire (JP)

58

Title Blue light alarm for a better morning : the effects of blue light ...

Documents