i-m-Walk

Interactive Multimedia Walking-Aware System

Meng-Chieh Yu

, Cheng-Chih Tsai

, Shih-Ta Liu

, Hao-Tien Chiang

, Ying-Chieh Tseng

Wei-Ting Chen

, Wan-Wei Teo

, Mike Y. Chen

, Ming-Sui Lee

1,2

and Yi-Ping Hung

1,2

Graduate Institute of Networking and Multimedia, National Taiwan University, Roosevelt Road, Taipei, Taiwan

Department of Computer Science and Information Engineering, National Taiwan University

Roosevelt Road, Taipei, Taiwan

Keywords: Smart Shoes, Walking Meditation, Visual Feedback, Slow Technology.

Abstract: i-m-Walk is a mobile application that uses pressure sensors in shoes to visualize phases of footsteps.

Through a mobile device, it will raise the awareness for the user to improve his walking behaviour. As an

example application in slow technology, we use i-m-Walk to help beginners to learn “walking meditation,” a

type of meditation where users aim to be as slow as possible in taking pace, and to land every footstep with

toes first. In our experiment, we asked 30 participants to learn walking meditation over a period of 5 days;

the experimental group used i-m-Walk from day 2 to day 4, and the control group did not use it. The results

showed that i-m-Walk could effectively assist beginners in slowing down the walking speed and while

decreasing the incorrect rate of pacing during walking meditation. To conclude, this study may serve as a

key in providing a mechanism to assist users to better understand his pacing and walking habits. In the

future, i-m-Walk could be used in other applications, such as walking rehabilitation.

1 INTRODUCTION

Walking and jogging is an integral part of our daily

lives in terms of transportation as well as exercise,

and it is a basic exercise can be done everywhere.

With the rapid growth of smartphones and the

development of human-computer interaction design,

many research projects have studied walking-related

interfaces through mobile phones. For example,

there is a research which evaluated the walking user

interfaces for mobile devices (Kane et al., 2008), and

proposed minimal attention user interfaces to

support ecologists in the field (Pascoe et al., 2000).

In addition, there are several walking-related

systems developed to help people in walking and

running. Nike uses touch sensor that attached to

users’ shoes to track the jogging information and

plays appropriate songs to user while jogging (Nike+,

2009). Adidas uses an accelerometer to detect user’s

pacing and heartbeats while jogging, and provides

the jogging information audibly, such as jogging

time, distance, and calories burned (miCoach, 2010).

Wii fit uses a balance board to detect user's centre of

gravity, and there are several games, such as yoga,

gymnastics, aerobics, and other balancing games

(Wii Fit, 2009). In addition, walking is a gentle, low-

impact exercise that can ease people into a higher

level of fitness and health. There are some research

uses walking as the rehabilitation exercises and an

essential exercise for elders (Femery et al., 2004).

Besides, the concept of “slow technology” is

proposed in recent years. Slow technology aims that

users should have more time to think and reflect

while learning, understanding and healing (Hallnäs

et al., 2001). The technology of ambient light and

biofeedback were some kinds of slow technology

and were widely used in the field of rehabilitation

and healing application now. Therefore, “walking

meditation” is one kind of meditation, and is also the

application of slow technology. Although many

research projects have focused on meditation,

showing benefits such as enhancing the

synchronization of neuronal excitation (Lutz et al,

2004) and increasing the concentration of antibodies

in blood after vaccination (Davidson, 2003), most

projects have focused on meditation while sitting.

There is little relevant research which uses

technology to help user learn the walking meditation.

Yu M., Tsai C., Liu S., Chiang H., Tseng Y., Chen W., Teo W., Chen M., Lee M. and Hung Y..

i-m-Walk - Interactive Multimedia Walking-Aware System.

DOI: 10.5220/0003118800170026

In Proceedings of the International Conference on Health Informatics (HEALTHINF-2011), pages 17-26

ISBN: 978-989-8425-34-8

 2011 SCITEPRESS (Science and Technology Publications, Lda.)

Figure 1: A participant is using i-m-Walk.

In this paper, in order to develop a system that

can help users to improve their walking habits. We

use the training of walking meditation as an example

application to evaluate the effectiveness of i-m-Walk.

Traditional training of walking meditation demands

on one-on-one instruction, and there is no

standardized evaluation. It is challenging for

beginners to self-learn walking meditation without

any feedback. We have designed experiments to test

the effect of walking awareness by using i-m-Walk

during walking meditation. Participants were asked

to do a 15-minute practice of walking meditation for

five consecutive days. During the experiment, the i-

m-Walk system will show real-time pace information

on the screen. We proposed two hypotheses: (a) i-m-

Walk could help beginners to walk slower during

walking meditation; (b) i-m-Walk could help users to

walk correctly while walking meditation.

This paper is structured as follows: The first

section deals with the introduction of walking

system. The second section of the article is a review

of walking detection and multimedia-assisted

walking applications. This is followed by some

introduction of walking meditation. The fourth

section describes the system design. After which

experimental design is presented. The results for the

various analysis are presented following each of

these descriptive sections. Finally, the discussion

and conclusion are presented and suggestions are

made on further research.

2 RELATED WORKS

In this section, we will discuss relevant literature on

the methods of walking detection, as well as other

cases where multimedia played a crucial part in

assisted walking applications.

The first concept of wearable computing and

smart clothing system included an intelligence cloth,

glasses, and an intelligence shoes, and the

intelligence shoes could detect the walking condition

(Mann, 1997). Then, in past decade, there were

many studies performed on this field, such as a study

where pressure and gyro sensors were used to detect

user’s feet posture, including heel-off, swing, and

heel-strike (Pappas et al., 2004). Moreover, a

research embedded pressure sensors in the shoes to

detect the walking cycle, and used vibrator to assist

users while walking (Watanabe et al., 2005). Besides,

there were many different methods on walking

detection, such as using bend sensor (Morris et al.,

2002), accelerometer (Crossan et al., 2005),

ultrasonic (Yeh et al., 2007), and computer vision

technology (Quek et al., 2008) to analyze pace.

However, with the development of ubiquitous

computing, there were many studies combined the

technology of multimedia, feedback, and walking

detector to assist people in different applications.

Drobny designed an intelligent shoe that can detect

the timing of pace, and play the music to help users

to learn ballroom dancing. While it detected missed

pace while dancing, it would show some warning

messages to the user. (Drobny et al., 2009).

Paradiso developed a system which can detect

dancers’ pace and applied them to interact with the

music (Paradiso, 2002). Mann developed a system

which can writes out the music on a timeline along

the ground, and each pace activates the next note in

the song, it can train the ability of musical tempo

and rhythm training for children (Mann, 2006).

Reynolds designed a system which used visual

information to adjust foot trajectory during the

swing phase of a footstep when stepping onto a

stationary target (Reynolds, 2005).

In the application of walking in psychological

field, Montoya used lighted target was used to load

onto left side and right side of walkway, and stroke

patients could follow the lighted target to carry on

their pace. The results pointed out that stroke patient

might effectively get help by using vision and

hearing as guidance (Montoya, 1994). Also, an

fMRI study on multimedia-assisted walking

experiment showed that increased activation during

visually guided self-generated ankle movements,

and proved that multimedia-assisted walking is

profound to people (Christensen et al., 2007).

In the application of pace analysis in healthcare

field, Noshadi developed a system which can detect

the walking stability of elderly and thus prevented

falling down. The system monitored walking

behaviours and used a fall risk estimation model to

predict the future risk of a potential fall (Noshadi et

al., 2008). Intiso used electromyography

biofeedback system to evaluate the effect of

HEALTHINF 2011 - International Conference on Health Informatics

biofeedback on stroke and rehabilitation patients,

and the results showed that biofeedback could help

patients to recover the swing phase of foot-drop after

training (Intiso et al., 1994). Moreover, in the

application of pace analysis in entertainment,

Nintendo published some walking-related games in

recent years. Personal Trainer – Walking could

detect users’ pace through accelerometer, and users

were encouraged to walk and get better scores

(Nintendo, 2009). Wii is a well-known game console,

and there had some interactions that should detect

user’s pace information, such as jogging and

dancing (Wii Fit, 2009).

However, none of related research in our survey

has a complete study to use multimedia-assisted

biofeedback and intelligent shoes to help users to

aware of their pace and improve it. In this study, a

complete research on integrating walking-aware

shoes and multimedia-assisted feedback for learning

walking meditation was proposed. In this paper, we

will describe the measurement of walking signal, the

walking-aware shoes, the biofeedback mechanisms,

the experimental design, and the experimental

results.

3 WALKING MEDITATION

The mechanism of meditation has many kinds and

different postures, such as meditation in standing,

sitting, lying down on back, and walking postures.

While the user is mediating with the posture of

walking, he would tend to feel less dull, tense, or

easily distracted in walking meditation. In this paper,

we focus on the meditation in the posture of walking,

which is also named walking meditation. Walking

meditation is a moving meditation which aligns the

feeling of body inside and outside. Also, it would

help people to focus and concentrate on his mind

and body. Furthermore, it can also deeply investigate

our knowledge and wisdom.

Figure 2: Six phases of a footstep in walking meditation

(Thera, 1998).

The principle of walking meditation aims to be as

slow as possible in taking pace, and landing each

pace with toes first. The people could focus on the

movement of his feet while walking. The movement

can be divided into six phases, from the status of

raising, lifting, pushing, lowering, stepping, to

pressing. People will be aware of the movement in

each stage while walking meditation (see Figure 2).

It is important to stay aware of the feet sensation.

Constant practicing of walking meditation is an

effective way to align the mind and body, and to

enhance the feeling of mindfulness. Mindfulness can

be defined as careful, open-hearted, choiceless,

present moment awareness. While feeling

mindfulness, people will slow down the footstep

while walking meditation. With long-term practice

of walking meditation, it benefits people through

increasing patience, enhancing attention,

overcoming drowsiness, and leading to healthy body

(Hanh et al., 2006).

The goal of this study is to help beginners to

aware of their mind and body, and to get healthier

after walking meditation. Therefore, i-m-Walk

system was developed to assist beginners in learning

the walking methods of walking meditation,

4 SYSTEM DESIGN

i-m-Walk includes a pair of intelligent shoes, an

ZigBee-to-Bluetooth relay, and a smartphone. Inside

the intelligent shoes, there are three force sensitive

resistor sensors fixed underneath each shoe insole,

and the sensing values will transfer wirelessly

through the relay to the smartphone. We develop the

system that can analyze the pace information, and

show the information through visual feedback. The

system is running on a HTC HD2 smartphone which

running Window Mobile 6.5 with a 4.3-inch LCD

screen. The overview of the system is shown in

Figure 3.

Figure 3: The flow chart of i-m-Walk.

i-m-Walk - Interactive Multimedia Walking-Aware System

4.1 i-m-Walk Architecture

The shoe module is based on Atmel's high-

performance, low-power 8-bit AVR ATMega328

microcontroller, and transmits sensing values

through a 2.4GHz XBee 1mW Chip Antenna

module wirelessly. The module size is 3.9 cm x 5.3

cm x 0.8 cm with an overall weight of 185g (see

Figure 4), including an 1800mAh Lithium battery

can continuous use for 24 hours. We kept the

hardware small and lightweight in order not to affect

users while walking.

Inside the shoes, we use three force sensitive

resistor sensors to detect the pressure distribution of

user’s feet while walking. The sensing area of each

pressure sensor is 0.5 inch in diameter. The

intelligent shoes would be used to detect user’s

walking speed and walking method of walking

meditation. According to the recommendations of

the orthopaedic surgery department in National

Taiwan University Hospital, we use three force

sensitive resistor sensors fixed underneath the shoe

insole. The surgeon recommended that there were

three main sustain areas located at structural bunion,

Tailor’s bunion, and heel, seperately. (see Figure 4).

Besides, the shoe module is put outside of the shoes

(see Figure 5). With a fully charged battery, the shoe

module can be used continuously run for 24 hours.

Figure 4: Shoe module. Right figure shows the micro-

controller and wireless module. Left figure shows one

insole embedded with three force sensitive resistor sensors.

Figure 5: Intelligent shoes. A sensing module is attached

onto the side of the shoe. The sensing values will transfer

wirelessly through the XBee module.

4.2 Walking Detection

According to different applications, there are many

sensing technologies and detection algorithm on the

field of walking detection (Pappas et al., 2004 &

Long et al., 2008). In our system, we use six

pressure sensors to sampling the pressure

distribution of user’s feet while walking. The sample

rate of our system is 30 times per second. In order to

detect whether the user lands each footstep with toes

first or not, the pressure distribution is divided into

two parts, front part and heel part. The sampling

value in front part is the average of two force

sensors which underneath at the position of

structural bunion and Tailor’s bunion. The sampling

value of heel part is a force sensor underneath at the

position of heel. Therefore, there are four sampling

values in our system to represent his walking status.

Then, we took an experiment to find out the most

appropriate threshold value which could categorize

whether the sampling value in these parts is pressed.

The experimental result showed that this method can

accurately detect the landing moment during the

weight of participants from 40 kg to 90 kg. Also,

this method can also detect that the user lands with

toes first or heel first. The definition of the

beginning of each gait cycle is in the moment when

the heel is lifted while walking. The end of the gait

cycle is in the moment while another foot’s heel is

rising while walking. Figure 6 shows an example of

our detection method. In this example, the system

detects that the user lifts his left foot in fifth second

because that the sensing value in heel part is less

than the threshold. Also, the system detects that the

user lifts his right foot in 10.7 second because that

the sensing value in heel part is less than the

threshold. Besides, if the sensing value in front part

is less than the threshold before front part, it means

that the user land this footstep with heel first.

Figure 6: Signal processing of walking signals. Blue line

indicates the sensing weight (kg) of heel, and green line

indicates the sensing weight of toes over time. Red line is

the threshold to activate the landing events. Gray block

represents which foot is landing.

HEALTHINF 2011 - International Conference on Health Informatics

4.3 USER INTERFACE

Multimedia assisted system can be effectively

applied in preventive medicine (Hu et al., 1994), and

it can also assist patients in walking easily (Femery

et al., 2004 & Woodbridge et al., 2009). In our

study, i-m-Walk is developed to assist beginners in

learning the walking methods of walking meditation.

The user interface of i-m-Walk includes three

components: warning message, footstep awareness,

and walking speed (see Figure 7). In this section, we

describe the user interface and the design principles

of our system.

4.3.1 Walking Aware

The main function of i-m-Walk is to reflect the

walking conditions on the mobile phone, and user

can aware of his walking conditions by watching the

context in real time. A pair of footmark is in the

center of the user interface, and it shows user’s

walking phases by using the movement of color

blocks. The color blocks show the center of gravity

in both feet separately, and the transparency will

change according to the volume of landing force.

Also, the color block will move top-down while the

user is landing with toes first. The color block will

move bottom-up while the user is landing with heel

first. Besides, while the user uses incorrect pace

while walking meditation, the color block will

change from green to red to remind the user to land

footstep with toes first in next footstep. In our study,

the correct walking method is defined that the user

lands every pace with toes first, and the color block

will show represent in green. On the contrary, while

the user lands footstep with heel first, the system

would recognize that he is using incorrect walking

method, and the color block would represent in red.

4.3.2 Walking Speed and Warning Message

During walking meditation, people should stabilize

his walking pace at a lower speed. i-m-Walk

provides the user interface that shows the walking

information in real time, and provides the

mechanism to remind user while he is walking too

fast. The walking speed is shown on the bottom of

the screen, and the speed is visualized as a

speedometer. In the speedometer, the indicator will

point to the corresponding walking speed such as the

speedometer for car. For example, if the indicator

points to the value “30”, it means that the user walks

thirty footsteps per three minutes. Therefore, the

speedometer also provides the function to remind the

user while he is walking too fast. According to the

experiment, we defined the lower-bound of the

walking speed as 40 footsteps per three minutes.

While the walking speed exceeds the speed, the

indicator will point to the red area, and the screen

will show a warning message “too fast” on the top of

the screen. The warning message would disappear

while the walking speed is less than 40 footsteps per

three minutes.

Figure 7: The user interface of i-m-Walk. The user

interface shows three components: warning message,

walking-aware, and walking speed. Left figure shows that

the user uses incorrect walking method on right foot. Right

figure shows that the walking speed is too fast (46

footsteps per three minutes), and the warning message

appears on the top of the screen.

5 EXPERIMENT DESIGN

Two experiments were designed to evaluate the

effects of walking-aware system during walking

meditation. The first experiment was a pilot study.

In this study, we evaluated the effect of visual

feedback which showed six sensing curves which

projected on the wall. The second experiment

evaluated the effects of i-m-Walk.

5.1 Pilot Study

Before the development of i-m-Walk, a pilot study

was held to test whether the visual feedback of

user’s walking information could help user to walk

better while walking meditation. Eight master

students volunteered to participate in this pilot study.

Participants’ average age is 26.3 (SD=0.52). All

participants have the experience of sitting meditation

before, but all of them do not have the experience of

walking meditation. There were four participants

(three male and one female) in experiment group

(with visual feedback), and four participants (four

male) in control group (without visual feedback).

Participants would take ten minutes each day and for

Warning

message

Walking

aware

Walking

speed

i-m-Walk - Interactive Multimedia Walking-Aware System

three consecutive days in this experiment. Before the

experiment, participants were taught the methods

and principles of walking meditation. The

experiment was a 4 × 2 between-participants

design. In the experiment group, participants were

asked to watch the walking information which

projected feet’s pressure distribution on the wall. In

the control group, participants were asked to walk

without watching the walking information. All

participants walked straight in the seminar room.

The results showed that there was a significant main

effect that experimental group had lower walking

speed than control group (p<0.05) during the three

days. Also, the median value of incorrect pace in

experimental group was less than control group, too.

The incorrect pace was defined in this study landing

footstep with heel first. As the results from the pilot

study, we concluded two preliminary conclusions: (a)

visual biofeedback could help beginners to slow

down the walking speed during walking meditation;

(b) Multimedia guidance could usefully help user to

aware of his pace during walking meditation, and

could decrease the number of incorrect pace.

However, we also observed a problem that

participant’s perspective would change over time

while walking, and it might influence the effect of

learning. Based on the results and recommends, we

developed a mobile application, i-m-Walk, to

visualize the walking information on a mobile

device, and designed a experiment to evaluate the

effect of it.

5.2 User Study

5.2.1 Participants

Thirty master and PhD students in the Department of

Computer Science volunteered to participate in this

experiment. Participants’ average age is 25.2

(SD=3.71). The results of questionnaires showed

that twenty-seven participants have the experience

of sitting meditation, and three participants do not.

Also, all participants do not have the experience of

walking meditation. 83.3% of the participants carry

mobile phone all the time, and 63.3% of the

participants have the experience of using

smartphone. There were fifteen participants (eleven

male and four female) in experiment group (with

visual feedback), and fifteen participants (eleven

male and four female) in control group (without

visual feedback). Because the feet size would differ

to each participant, we prepared two pair os shoes

with different sizes for participants to choose a

comfortable one.

5.2.2 Location

Meditating in a quiet and enclosed area would be

easier to bring mind inward into ourselves and may

reach in calm and peace situation. In this experiment,

we selected a corridor in the faculty building as the

experimental place for walking meditation. The

corridor is a public place at an enclosed area and few

people would conduct their daily activities like

standing, walking, and interacting with one another

there. The surrounding of corridor is quiet and

comfortable for participants to reach their mind in

calm. The length of the corridor is thirty meters, and

the width is three meters. The temperature is 21~23

Celsius degree.

5.2.3 Procedure and Analysis

Before the experiment, participants were asked to

walk alone the corridor in casual walking, and we

recorded the walking speed. We would like to ensure

that all participants are under the same condition

before the training of walking meditation. The result

showed that the average walking speed is 95.97

footsteps per minute (SD = 4.67). Then, we taught

the methods of walking meditation to all

participants. The guideline of the walking meditation

which we provided to the participants was as

follows: “Walking meditation is a way to align the

feeling inside and outside of the body. You should

focus on the movement of footstep while walking.

Each footstep would include six phases, from the

phase of raising, lifting, pushing, lowering, stepping,

to pressing. Besides, you have to land every footstep

with toes first and then slowly land your heel down.

During walking meditation, you should stabilize

your walking pace at a lower speed as possible. You

have to relax your body from head to toes.”

Table 1: Experimental procedure: ● means that

participants were asked to use i-m-Walk and ○ means that

participant do not use i-m-Walk during walking

meditation.

DAY

1

DAY

2

DAY

3

DAY

4

DAY

5

Experimental

Group

○ ● ● ● ○

Control Group ○ ○ ○ ○ ○

The experiment was a 15 × 2 between-

participants design. Participants would take fifteen

minutes each day and for five consecutive days in

this experiment. Table 1 shows the procedure of this

HEALTHINF 2011 - International Conference on Health Informatics

experiment. In the experimental group, participants

were asked to use i-m-Walk from day 2 to day 4. In

the control group, participants were asked to walk

without any feedback during walking meditation.

While learning of walking mediation, all

participants were asked to walk clockwise around

the corridor and hold the smartphone. In control

group, there was no visual feedback on smartphone

although they still needed to hold. In experimental

group, participants were informed that they can

decide not to see the visual feedback while they did

not need it. The participants in experimental group

were asked to complete a questionnaire after each

task from day 2 to day 4. Besides, we asked all

participants the feeling and impression after the

experiment in day 5. However, all participants could

write down any recommends and feelings after the

experiment, and we will discuss the issues in the

discussion section.

5.2.4 Results

We analyzed the walking speed and incorrect pace

both in experimental group and control group. In the

results of the walking speed, figure 8 shows the

average stride time on experimental group and

control group from day 1 to day 5. In day 1 and day

5, all participants learned walking meditation

without using i-m-Walk. Following t-tests revealed

significant difference (p < 0.005) that the

experimental group had longer stride time than

control group from day 2 to day 4. In experimental

group, the average of stride time increased from 4.5

seconds in day 1 to 10.9 seconds in day 5. In control

group, the average of stride time increased from 3.2

seconds (day 1) to 5.1 seconds (day 5). The results

showed that the participants in experimental group

had significant main effect (p < .005) in slowing

down the walking speed after the learning of

walking meditation. On the contrary, the participants

in control group had no significant main effect (p

> .1) in slowing down the walking speed after the

learning of walking meditation. Therefore, the

results showed that i-m-Walk could help participants

to slow down the walking speed during walking

meditation.

In our experiment, the definition of correct

walking method is that users should land every

footstep with toes first during walking meditation. If

participants landed footstep with heel first, it was an

incorrect pace. Figure 9 shows the median values of

total incorrect pace in 15 minute learning of walking

meditation on experimental group and control group

from day 1 to day 5. In experimental group, the

median value of incorrect pace decreased from eight

in day 1 to one in day 5, and it decreased over day.

In control group, the median value of the incorrect

pace decreased from seven pace in day 1 to five ace

in day 5, but the incorrect pace decreased only in the

first three days. The results showed that i-m-Walk

could effectively reduce incorrect pace during

walking meditation.

Figure 8: The comparison of average stride time on

experimental group and control group from day 1 to day 5.

Error bars show ±1 SE.

Figure 9: The comparison of the median of incorrect pace

numbers on experimental group and control group from

day 1 to day 5.

The experimental group was asked to complete a

questionnaire after using i-m-Walk system from day

2 to day 4. There are two questions in the

questionnaires, and the content was the same in each

day. Figure 10 shows the results of questionnaires.

We asked two questions: (1) what is the degree of i-

m-Walk to help you to aware of the walking pace? (2)

what is the degree of i-m-Walk to help you tslow

down the walking speed? There were five options,

including “1: serious interference”, “2: a little

4,5

9,2

10,8

10.9

3,2

3,8

4,3

5,1

day1 day2 day3 day4 day5

Time (sec)

ExperimentalGroup

ControlGroup

day1 day2 day3 day4 day5

Wrong Pace (footstep)

Experimental

Group

i-m-Walk - Interactive Multimedia Walking-Aware System

interference”, “3: no interference and no help”, “4: a

little help”, and “5: very helpful”, and participants

should fill the answer. The results of questionnaires

showed that all participants in experimental group

gave positive feedback both in question 1 and

question 2, and almost all participants in

experimental group liked the system.

Figure 10: The results of questionnaires which were filled

by experimental group from day 2 to day 4. The red line

means the baseline of the satisfaction. While the value

above the baseline means that the participant agrees to the

question. Error bars show ±1 SE.

6 DISCUSSION

The aim of this section is to summarize, analyze and

discuss the results of this study and give guidelines

for the future development of applications.

6.1 User Interface

The user interface of i-m-Walk provides the walking

information, including walking speed, incorrect pace,

and the center of feet. The experimental results

showed that i-m-Walk could help beginners to

decrease the walking speed while walking

meditation. There is a participant’s comment from

experimental group:

User E6 in day 3: “I always walked fast, but when I

saw the dashboard and the warning message “too

fast” on the screen, it was helpful to remind me to

slow down the walking speed.

The experimental results also showed that i-m-

Walk could effectively reduce the rate of incorrect

pace for beginners while walking meditation. One of

the participants from experimental group said that:

User E 1 in day2: “While the color block changed

the color from green to red, I knew that I used

incorrect pace immediately. Then, I would pay

attention on my pace deliberately in the next footstep.

The mechanism of remind function likes a

personal couch to remind user while using incorrect

pace in walking meditation. We concluded some

design principles of the user interface: (a) i-m-Walk

used the form of dashboard to represent the walking

speed. The value of walking speed is easy to watch,

and user might aware of the change of walking

speed while he slowed down or speeded up; (b) i-m-

Walk provided additional alarm mechanism, a

warning message “too fast”, while walking too fast.

The mechanism could remind user when he is

distraction; (c) i-m-Walk could effectively reduce the

rate of incorrect pace for beginners. Besides, two

participants said that they always forgot to breathe

while concentrating on the user interface. It is

possible to provide the guidance of breathing rhythm

with the rhythm of walking during walking

meditation.

6.2 Human Perception

Vision, sound, smell, taste and touch are five main

perceptual modalities for human beings. The most

use in human-computer interaction is visual

modality and audio modality now. There was a

comment from a participant from experimental

group:

User E3 in day 2: “If I can listen to my pace during

walking meditation, I do not need to hold the

smartphone”.

In cross-modal research, visual modality is

always considered superior than auditory modality

in spatial domain. In our case, i-m-Walk needs to

show the walking information, walking speed and

incorrect pace in the same time. Therefore, i-m-Walk

uses visual feedback as the user interface. The

advantage of visual feedback is that users could

choice to watch the information or not, but the

shortcoming is that users cannot receive the

information while they do not watch it. Therefore, it

is possible to provide more interaction methods to

remind users, such as tactile perception and acoustic

perception. On the other hand, the mechanisms of

multimedia-assisted feedback might attract user’s

attention in some case. Too many inappropriate and

redundant events might disturb user. In our system,

i-m-Walk provides visual feedback all the time

during walking meditation because we do not know

whether the user needs the information or not.

Therefore, we informed all participants that they

could choice not to watch the screen while they

could aware of their pace well. By this way, it could

4,07

4,29

4,07 4,07

4,23 4,23

0,5

1,5

2,5

3,5

4,5

day 2 day 3 day 4

i-m-Walk can

help you increase

the aware

i-m-Walk can help

you decrease the

walking speed

HEALTHINF 2011 - International Conference on Health Informatics

minimize the interference while walking meditation.

6.3 Beginner vs. Master

In recent years, the concept of “slow technology”

was applied in many mediate systems. The design

philosophy of “slow technology” is that we should

use slowness in learning, understanding and

presence to give people time to think and reflect. In

our study, walking meditation is a form of slow

technology. There are two main parts in walking

meditation, inside condition and outside condition.

The inside condition means the meditation of mind

and the outside condition means the meditation of

walking posture. All participants were beginners in

our experiment because we would like to focus on

the training of the walking posture first. The

difference between beginner and master in walking

meditation is that the beginner does not familiar to

walking meditation and needs to pay more attention

on the control of walking posture; on the contrary,

the master familiar to it and could pay more

attention to align the mind and walking posture in

the same time. Walking meditation is a way to align

the feeling inside and outside of the body. The

beginner should familiar the walking posture before

the spiritual development. In this paper, the goal of

our experiment is to evaluate the learning effects of

i-m-Walk system. The experimental results showed

that experimental group could slow down the

walking speed and decrease incorrect pace after five

days. Six participants in experimental group felt that

the experimental time in day four was short than first

day although the experimental time was the same.

However, there was no such comment from the

participants in control group. The results showed

that i-m-Walk could help user to train the walking

posture of walking meditation.

6.4 Reaction Time

Reaction time is an important issue in human-

computer interaction design. If the reaction time is

too long, it is hard for user to use it. According to

the observation, the delay time of i-m-Walk is about

0.2 second. However, the delay time do not affect

the user because the application in this experiment

does not need fast reaction time. The average pace

speed is 10.9 seconds in experiment group in day

five. The results of questionnaires also showed that

participants felt that the visual feedback could reflect

their walking status immediately.

7 CONCLUSIONS AND FUTURE

WORK

In this paper, we present a mobile application that

uses pressure sensors in shoes to visualize phases of

footsteps on a mobile device in order to raise the

awareness for the user´s walking behaviour and to

help him to improve it. Our study shows that i-m-

Walk can effectively assist beginners in slowing

down the footstep frequency and eliminating the

error rate of pace during walking meditation.

Therefore, i-m-Walk can be used in other

applications, such as walking rehabilitation, fitness,

and entertainment applications.

Despite the encouraging results of this study as

to the positive effect of i-m-Walk, future research is

required in a number of directions. First, the

experiment in this study showed that i-m-Walk could

help beginners to aware of their footsteps, but we

did not know the learning effect in the long-term.

How long a beginner can familiar to the essence of

walking meditation is an interesting issues. In the

future, we will record user’s learning status day by

day, and to find out the learning curves of walking

meditation. Second, i-m-Walk only used the

mechanism of visual feedback to assist users, and it

did not include the mechanism of auditory feedback

and tactile feedback. In the future, we would like to

compare the effect of visual feedback, auditory

feedback, and tactile feedback while walking. Third,

breathing method is important while walking,

jogging, and running. We are developing system

named i-m-Treadmill, which investigates the

relationship between jogging and breathing. Finally,

we will develop a system which integrates the

technique of breath detection, walking detection, and

multimedia-assisted biofeedback to develop a health

care system, and toward to the ideal of preventive

medicine.

ACKNOWLEDGEMENTS

This work was supported in part by the National

Science Council, Taiwan, under grants NSC 98-

2221-E-002-128-MY3, and by the Technology

Development Program for Academia, Ministry of

Economic Affairs, Taiwan, under grant 98-EC-17-

A-19-S2-0133. And thanks to Footwear &

Recreation Technology Research Institute, Taipei

National University of the Art, and Dharma Drum

Buddhist College.

i-m-Walk - Interactive Multimedia Walking-Aware System

REFERENCES

adidas, 2010. miCoach, Retrieved March 5, 2010, from

www.micoach.com/

Christensen, M. S., Lundbye-Jensen, J., Petersen, N.,

Geertsen, S. S., Paulson, O. B., and Nielsen, J. B.,

2007. Watching Your Foot Move--An fMRI Study of

Visuomotor Interactions during Foot Movement.

Journal of Cereb Cortex, 17 (8), 1906-1917.

Crossan, A., Murray-Smith, R., Brewster, S., Kelly, J., and

Musizza, B., 2005. Gait phase effects in mobile

interaction, In Proc. CHI '05 extended abstracts on

Human factors in computing systems, 1312-1315.

Davidson, R. J., 2003. Alterations in brain and immune

function produced by mindfulness meditation.

Psychosom Med., 65, 564-570.

Drobny, D., Weiss, M., and Borchers, J., 2009. Saltate! –

A Sensor-Based System to Support Dance Beginners.

In Proc. CHI '09 Extended Abstracts on Human

Factors in Computing Systems, 3943-3948.

Femery, V. G., Moretto, P. G., Hespel, J.-M. G., Thévenon,

A., and Lensel, G., 2004. A real-time plantar pressure

feedback device for foot unloading. Journal of Arch

Phys Med Rehabi, 85(10), 1724-1728.

Hanh, T. N., Nquyen, A.-H., 2006. Walking Meditation.

Sounds True, Incorporated.

Hu, M. H., and Woollacott, M. H., 1994. Multisensory

training of standing balance in older adults: I. Postural

stability and one-leg stance balance. Journal of

Gerontology: MEDICAL SCIENCES 49(2), 52-61.

Hallnäs, L. and Redström, J., 2001. Slow Technology:

Designing for Reflection. Personal and Ubiquitous

Computing, 5(3), 201-212.

Intiso, D., Santilli, V., Grasso, M.G., Rossi, R., and

Caruso, I., 1994. Rehabilitation of walking with

electromyographic biofeedback in foot-drop after

stroke. Journal of Stroke, 25(6), 1189-1192.

Ikemoto, L., Arikan, O., and Forsyth, D., 2006. Knowing

when to put your foot down. In Proc. Interactive 3D

graphics and games 06’, 49-53.

Kane, S. K., Wobbrock, J. O., and Smith, I. E., 2008.

Getting off the treadmill: evaluating walking user

interfaces for mobile devices in public spaces. In Proc.

MobileHCI '08, 109-118.

Kong, K. and Tomizuka, M., 2008. Smooth and

Continuous Human gait Phase Detection based on foot

pressure patterns. In Proc. ICRA ’08, 3678-3683.

Lutz, A., 2004. Long-term meditators self-induce high-

amplitude gamma synchrony during mental practice.

PNAS, 101, 16369-16373.

Mann, S., 1997. Smart clothing: The wearable computer

and WearCam. Journal of Personal Technologies, 1(1),

21-27

Mann, S., 2006. The andante phone: a musical instrument

that you play by simply walking. In Proc. ACM

Multimedia 14th, 181-184.

Montoya, R., Dupui, P. H., Pagès, B., and Bessou, P.,

1994. Step-length biofeedback device for walk

rehabilitation. Journal of Medical and Biological

Engineering and Computing, 32(4), 416-420.

Morris, S. J., and Paradiso, J. A., 2002. Shoe-integrated

sensor system for wireless gait analysis and real-time

feedback. In Proc. Joint IEEE EMBS (Engineering in

Medicine and Biology Society) and BMES (the

Biomedical Engineering Society) 2nd, 2468-2469.

Nike, 2009. Nike+, Retrieved October 26, 2009, from

www.nikeplus.com/

Nintendo, 2009. Personal Trainer - Walking, Retrieved

November 26, 2009, from

http://www.personaltrainerwalking.com/

Nintendo, 2009. Wii Fit Plus, Retrieved January 8, 2010,

from http://www.nintendo.co.jp/wii/rfpj/index.html

Noshadi, H., Ahmadian, S., Dabiri, F., Nahapetian, A.,

Stathopoulus, T., Batalin, M., Kaiser, W., and

Sarrafzadeh, M., 2008. Smart Shoe for Balance, Fall

Risk Assessment and Applications in Wireless Health.

In Proc. Microsoft eScience Workshop ‘08.

Paradiso, J., 2002. FootNotes: Personal Reflections on the

Development of Instrumented Dance Shoes and their

Musical Applications. In Quinz, E., ed., Digital

Performance, Anomalie, digital_arts, 2, 34 - 49.

Pascoe, J., Ryan, N. and Morse, D., 2000. Using while

moving: HCI issues in fieldwork environments. ACM

Transactions on Computer-Human Interaction, 7 (3),

417- 437.

Pappas, P. I., Keller, T., Mangold, S., Popovic, M. R.,

Dietz, V., and Morari, M., 2004. A reliable, insole-

embedded gait phase detection sensor for FES-assisted

walking. Journal of IEEE Sensors, 4 (2), 268-274.

Quek, F., Ehrich, R., and Lockhart, T., 2008. As go the

feet...: on the estimation of attentional focus from

stance. In Proc. ICMI 10th, 97-104.

Reynolds, R.F., Day, B.L., 2005. Visual guidance of the

human foot during a step. Journal of Physiology, 569

(2), 677-684.

Thera, S., 1998. The first step' to Insight Meditation.

Buddhist Cultural Centre.

Watanabe, J., Ando, H., and Maeda, T., 2005. Shoe-

shaped Interface for Inducing a Walking Cycle. In

Proc. ICAT 15th, 30-34.

Woodbridge, J., Nahapetian, A., Noshadi, H., Kaiser, W.

and Sarrafzadeh, M., 2009. Wireless Health and the

Smart Phone Conundrum. HCMDSS/ MDPnP.

Yeh, S. Y., Wu, C. I., Chu, H. H., and Hsu, Y. J., 2007.

GETA sandals: a footstep location tracking system.

Personal and Ubiquitous Computing, 11(6): 451-463.

HEALTHINF 2011 - International Conference on Health Informatics