Automatic Attendance Control System based on BLE Technology

Miran Bori

c, Ana Fern

andez Vilas and Rebeca P. D

ıaz Redondo

Information & Computing Lab., AtlantTIC Research Center, School of Telecommunications Engineering,

University of Vigo, 36210 Vigo, Spain

Keywords:

Bluetooth Low Energy, iBeacons, Experiment, Data Collection, Broadcasting.

Abstract:

As an emerging low power wireless technology, Bluetooth Low Energy lately has become an essential part

of the Internet of Things world. Its pervasiveness has given the possibility for many different, otherwise

manually driven, concepts to work in a different, enhanced, and more automated fashion. In this paper, we

present an automatic attendance control system based on the interaction between a modiﬁed BLE iBeacon

mechanism and users’ smart devices. As such, the system can recognize users according to predeﬁned proﬁles

and eventually broadcast valuable information that corresponds to recognized proﬁles. The proof of concept

shows that this is a promising approach regarding automating speciﬁc tasks, recognizing users and interacting

with them less complexly.

1 INTRODUCTION

Since the release of Apple’s iBeacon protocol back

in 2013 (Newman, 2014), Bluetooth Low Energy be-

came a dominant technology for many projects in the

Internet of Things and human life. From small user

appliances to big industry wireless solutions, every-

thing is trying to be connected. With an estimated 21

billion devices possibly connected by the year 2020,

data generating a massive amount of data becomes a

usual daily routine. With that said, businesses can no

longer ignore such growth and should adjust to this

revolution swiftly. Along with the IoT, a new concept

called ”The Physical Web” embedded itself into BLE

technology and lately is on the rise. Such idea allows

us to connect the physical and virtual world with each

other. Better said, it enables us to interact with the

objects that surround us, such as user appliance de-

vices. The product of such interaction is sharing and

processing of different data to gain a better knowledge

and to facilitate everyday activities.

Surely, with the smart environment and the IoT,

the door to different technological approaches, where

devices share essential information is wide open. We

do see that BLE is taking over the world, mainly in

the consumer device area, where the whole ecosystem

proceeds to grow continually. Especially with the lat-

est BLE concept called Bluetooth proximity beacons

often referred to as the Beacons (ibeacons), where LE

peripherals advertise themselves to the nearby devices

by using advertisement packets (Martin et al., 2014).

An iBeacon represents an electromagnetic sig-

nal that can be processed by every BLE enabled de-

vice. In more details, iBeacons are small transmit-

ters that advertise a speciﬁc BLE payload that is rep-

resented by particular identiﬁers: UUID (Universally

Unique Identiﬁer) 128-bit value, Major value (16-bit

unsigned integer), and Minor value (16-bit unsigned

integer). The last two values are optional and are in-

cluded in case of more hierarchical system implemen-

tations. As already said, an iBeacon signal can repre-

sent a speciﬁc object or process in some deﬁned en-

vironment. Later, when received by a BLE enabled

device, that signal triggers some action that means

something to the user or some other object.

Nowadays, numerous beacon maker companies

exist, and every company has something different to

offer. Their products differ in various sizes, function-

alities, but the majority of them are static. In other

words, a single beacon device can only broadcast one

iBeacon combination of UUID, Major, and Minor val-

ues at the same time, and any further classiﬁcation is

handled by the application that is usually installed on

the users’ smartphones. If we want to broadcast dif-

ferent type of data, we need another beacon device

with separate iBeacon combination. Developing an

app became a primary focus and the central point of

all the processing, leaving the iBeacon hardware as a

static device without any advanced functionality. As

already mentioned, many solutions exist on this sub-

ject, but most of the processing is done on the appli-

cation side. In most cases, it became mandatory for

Bori

c, M., Vilas, A. and Redondo, R.

Automatic Attendance Control System based on BLE Technology.

DOI: 10.5220/0006830202890295

In Proceedings of the 15th International Joint Conference on e-Business and Telecommunications (ICETE 2018) - Volume 1: DCNET, ICE-B, OPTICS, SIGMAP and WINSYS, pages 289-295

ISBN: 978-989-758-319-3

289

the user to have an application installed on his smart-

phone. Let us not forget that the beacon device ini-

tially can not collect BLE signals since it is a one-way

broadcasting technology.

To overcome such a static method of broadcasting

iBeacon BLE signal, in our previous work we have

proposed a dynamic solution where a modiﬁed BLE

beacon device can simultaneously transmit and re-

ceive different signals according to various user pro-

ﬁles (Boric et al., 2018), as well as the implementa-

tion part and the challenges (M. Boric et al., 2018).

By modiﬁed beacon, we refer to a device that in the

same moment can work in a scanning mode, and

an iBeacon mode (advertising speciﬁc information to

corresponding users) (Townsend et al., 2014). Ad-

ditionally, these two modes complement each other,

what allows the device to broadcast information to a

speciﬁc group of users that share the same proﬁle.

The proﬁle can be created manually by predeﬁned

taxonomy, or automatically by users’ movement post

processing and detailed analytics. For this simultane-

ous broadcasting of information to different proﬁles,

the device needs to have a particular mechanism and

the background infrastructure that allows the proﬁle

recognition and the broadcast. Now, we are expand-

ing that system with real-world examples to show its

usage and beneﬁt for different purposes.

The recognition concept above could contribute

to solving the problem of manual presence detection

through automatically collecting users’ smartphones

BLE signals in any scenario environment where such

detection is required. To show the proof of concept we

utilize the mentioned system for collecting students’

lecture attendance in the academic environment. Our

scenario experiment took place at the Faculty of En-

gineering and Telecommunications of Vigo in Spain.

We have planted one beacon device in the hallway,

for testing the dynamic beacon broadcasting, and an-

other device in one of the classrooms of the Faculty

for collecting student attendance.

The experiment was conducted in a one-month pe-

riod, and we present the details in the following sec-

tions. In this work, we show how our dynamic sys-

tem can be used in the academic environment to au-

tomatically collect students lecture attendance when

students are in beacon proximity, without any appli-

cation installed on the user side. However, with the

application installed on the user side, a system may

eventually broadcast information to him, e.g., dif-

ferent events, classes timetable, homework, faculty

changes, and so on.

The remainder of this paper we divide into several

sections. Section 2 gives the related work and what

was our primary motivation. Part 3 explains the sys-

tem architecture and how we use it to for proposed

system. In section 4 we present the implementation

part with the results, wherein section 5 we conclude

the paper.

2 RELATED WORK

Researchers have been advancing in the Bluetooth

Low Energy ﬁeld lately. This progress gave some

great examples of the use of such technology in to-

day’s world, from retail business (Shende et al., 2017)

and giving a better experience to the customers, to

monitoring different processes in industry or health

(Srinivasan et al., 2016) (Komai et al., 2016). When

speaking of beacon involvement in the academic en-

vironment, few projects have been proposed and de-

ployed. Researchers in (S. Barapatre et al., 2017)

have proposed the smart college management system

with the ability of automatic student class attendance.

However, for the system to be able to recognize a par-

ticular student, a student has to use an Android app

and to mark his attendance during a particular class.

Similarly, authors in (Apoorv and Mathur, 2016) have

built a system that facilitates the teachers to collect

student attendances. For this purpose, the Android ap-

plication communicates with the BLE beacon planted

in the classroom and collects the sensor data, from ID

cards, that matches the student’s attendance. How-

ever, to collect attendance, a teacher has to collect all

the BLE signals manually through the app.

Furthermore, authors in (Saraswat and Garg,

2016) are using beacons for faculty administrative

tasks and interacting with the students by sending

them Web links and the corresponding notiﬁcations.

Another feature of this system is the automatic stu-

dent attendance collection, where students according

to their stay in the classroom, during particular lec-

ture, get recorded by the system. That said, when the

speciﬁc student is near beacon a clock time starts in

the application background automatically and accord-

ing to its proximity (near, immediate), local clock-

time increases. However, the majority of the work is

done on the application side (student’s smartphone).

In our case, a dynamic beacon is the focal point of

all the processing that includes sending information

to the students, collecting their attendance, analyzing

their building usage and giving an insight to the fac-

ulty personnel.

Also, the reason for electing BLE technology is

because of its low power consumption, comparing to

Wi-Fi for instance (Putra et al., 2017). Also, Wi-Fi

technology asks for previously established connec-

tion between two hosts which is not convenient in our

WINSYS 2018 - International Conference on Wireless Networks and Mobile Systems

290

case. In case of the Wi-Fi direct and classic Blue-

tooth, they ask for secure pairing to unknown devices

(Trifunovic et al., 2011).

3 SYSTEM ARCHITECTURE

The idea is to have a unique system that recognizes

users while in a proximity of a beacon and triggers

some actions based on that recognition. The user’s

smart device can be used to monitor the proximity lo-

cation of a user in the beacon vicinity. Based on this

location information, a user’s presence record will

get updated automatically on the system. For bet-

ter understanding, we will take the example of stu-

dents lecture attendance. Usually, when students en-

ter a speciﬁc classroom, a professor manually checks

the students’ participation, either on a piece of pa-

per or using some application. This kind of work is

time-consuming and introduces complexity in further

school analysis, e.g., If the school board wants to get

better student/classes/subject insight, everything has

to be done manually. However, by using BLE in com-

bination with the student’s smartphone, or some other

BLE enabled device, this work can be automatized

and can save much time for the future faculty analy-

sis.

For better system explanation, we give the con-

cept in Figure 1, where the beacon device works in

two modes, the discovery, and the advertising mode

that uses the iBeacon technology. The discovery

mode, as already mentioned, is in a constant search

for the nearby devices (BLE devices) and the adver-

tising mode optionally sends some relevant interest-

based data to the recognized users. When a partic-

ular user enters a classroom (broadcasting area), his

smartphone BLE MAC address gets recognized by

the classroom beacon device that works in a discov-

ery (passive scanning) mode. For the user to be iden-

tiﬁed, it is not obligatory to have an installed applica-

tion on his smartphone, or even connect to the system

since the user is recognized automatically. Next, the

beacon device sends the user’s BLE MAC address to

the database through the back-end infrastructure. The

back-end can be based on any connection technology,

or even implemented on the same device where the

beacon functionality is installed. In our case, it is in-

stalled on one of the Raspberries and connected via

secured Wi-Fi connection. To send some information

to the recognized user, we need to use the advertising

mode of the beacon device (iBeacon mode). How-

ever, this step is optional and asks for an application

on the user’s side.

The database is nothing else but a student reposi-

Figure 1: Dynamic beacon functionality.

tory where all his relevant data is saved (Student year,

subjects, classroom he attended and time, personal

info, and so on.). Personal user information can be in-

serted through the faculty information system, while

all the other information (student movement activity)

can be collected from the database. Also, the database

contains a student attendance report of his every atten-

dance, for which the beacon device is used in the ﬁrst

place. In that case, the system would suffer from false

attendance inputs. This presence is not marked at the

same moment when the student enters the classroom

but after a speciﬁc period. This period depends on

how the faculty organizes and schedules their classes,

and also how does it handle students attendance. For

example, if the faculty agrees for the lessons to be

held for 2 hours than the beacon device can be conﬁg-

ured in that manner. A minimum number of student

acknowledgments by the system, during the lecture

time, is ﬂexible and conﬁgurable, and it is a subject of

a faculty policy, as mentioned before. If the student

is present in the classroom during the whole lecture

period (e.g., two hours), the system marks his atten-

dance as the database record.

Nevertheless, we have to consider a problem of

the signal overlapping, especially if the beaconing ar-

eas are a close one to another (two classrooms, of-

ﬁces, and so on). For example, if two students are

attending classes in the different classrooms, and the

classrooms are close as mentioned, it is possible for

the signal to overlap. This behavior would result

in collecting false data. In that case, the system

would not correctly process the inputs, and the out-

put would not be trustworthy. Even with the excellent

and planned site survey, we have to take into account

Automatic Attendance Control System based on BLE Technology

291

a certain amount of false positives. First, a precise

positioning of the beacon device and the proper site

survey of the classroom need to be done. Second, an

additional mechanism to prevent such behavior needs

to be introduced. In our case, that is a door scanner

positioned correctly at the classroom entrance, so that

the system can scan students when entering the room.

For that matter, we can use the infrared technology or

something similar and connect it and conﬁgure it to

the Raspberry device, dynamic iBeacon in our case.

The beacon device can work simultaneously for many

different proﬁles. In other words, it can recognize dif-

ferent proﬁles, if proﬁle recognition is implemented

into the database. In this case, all we need is one bea-

con device per every classroom. That said, we lower

the broadcasting congestion and network latency that

represent the problem for most of the IEEE 802 net-

works. The beacon device does not have to be a part

of the school infrastructure (school Wi-Fi, Ethernet,

and so on).

Figure 2: Beacon device architecture.

Although, every beacon device has to connect to

the central beacon repository. In our case, this repos-

itory is the database. As mentioned before, for the

beacon to function dynamically, constant connection

to DB is necessary. This way, the beacon advertiser

is changing his beacon combination ID, depending

on proﬁle entrance and, if needed, disseminates the

corresponding information to the users. Every proﬁle

is represented by a different combination ID (UUID,

Major, and Minor number). The most critical piece

of information for the system to be functional is the

user’s device MAC address with which the system

recognizes a particular user and his proﬁle. A user de-

vice does not have to be his private mobile phone ex-

plicitly, but any piece of hardware that supports BLE

(BLE token, smart card, and so on). However, if we

want to utilize a full system capacity, advertising in-

formation to the students, using smartphones repre-

sents a better approach. For the system to work ap-

propriately, student phone MAC addresses need to

be inserted into the database before the student at-

tendance collection. Achieving beacon to operate in

a dynamic mode is not an easy task. Usually, bea-

cons work in a static mode where every beacon sig-

nal identiﬁer (UUID, Major, and Minor values) cor-

responds to only one information, and they were ini-

tially imagined as such. However, having many bea-

con devices in a single area brings network latency

and congestion, as mentioned before. To overcome

that problem, we offer a solution where beacon de-

vice recognizes the users dynamically when near the

discovery/broadcasting area. Dynamically means that

the beacon can serve and acknowledge many user pro-

ﬁles simultaneously without degrading its functional-

ity (see Figure 2). In our case, we use such solution

for automatic student attendance, as mentioned ear-

lier. Additionally, if positioned correctly (along the

faculty area), the beacon system can collect all the

faculty users on a regular basis and save them into the

database. That said, this information could be valu-

able for building managers for better space analytics,

building management, on so on. If so, the faculty

board can use this insight for different future planning

regarding their assets and faculty management.

4 IMPLEMENTATION AND THE

RESULTS

As already mentioned, a single beacon device con-

tains two major parts; the discovery (scanning) and

the advertising part (iBeacon) where every element

works separately and includes its methodology. That

said, for the beacon device we use a Raspberry Pi

small computer and corresponding BLE USB module

devices that play a vital role in the dynamic broadcast-

ing. Beacon device is based on Raspberry Pi 3 Model

B device along with the Bluetooth 4.0 CSR8510 USB

Module. Reason for this is because the Raspberry

device is Linux based and in that manner, it is very

ﬂexible and scalable for this kind of solution. One

BLE USB module device is used for the discovery

part, while the other for advertising part (see Figure

2). The beacon device is connected to the database,

”Student Proﬁle DB” in our case. We have to mention

that the database can be implemented separately from

the beacon device on a different device, or it can be in-

stalled on the same device where the beacon resides.

This implementation is possible due to the ﬂexibility

of the Raspberry Pi Linux operating system. ”Stu-

dent Proﬁle DB” is based on the MySQL open-source

relational database management system. It is worth

WINSYS 2018 - International Conference on Wireless Networks and Mobile Systems

292

mentioning that detailed database structure is not pre-

sented since it was not relevant for the paper. In our

case, a single dynamic beacon device is planted in the

middle of the ceiling of one of the classrooms, for the

uniform distribution of signal, and it is activated to

work in a discovery mode (see Figure 3).

Figure 3: System architecture.

To evaluate the performance of the proposed

system, we have conducted an experiment on the

premises of the Faculty of Engineering and Telecom-

munications of Vigo in Spain. To show that single

beacon device is capable of acquiring different user

proﬁles, we planted it in one part of the faculty area

and gathered information during a short period. As

already mentioned before, user proﬁles had to be im-

plemented, according to some taxonomy and inserted

into the ”Student Proﬁle” database. We were able to

recognize and collect four users simultaneously what

can be seen in Figure 4. From the Figure 4 we can

see that the device after recognizing speciﬁc proﬁle

(user), by scanning his phone’s MAC address, it acti-

vates the script for some information distribution. The

device broadcasts an iBeacon combination ID (UUID,

Major, and Minor values) that corresponds to a partic-

ular user proﬁle. An example of the iBeacon combi-

nation and how the BLE advertising device (hci1) is

Figure 4: Simultaneously recognized users.

activated before sending the corresponding combina-

tion ID can be seen from the Figure 5. Let us not

forget that for the notiﬁcation part, an application has

to be developed and installed on the users’ smart de-

vices.

Figure 5: Combination ID script activation.

To acquire student attendance, we planted another

beacon device in one of the classrooms for precisely

one month and monitored the lecture that students

had on Monday’s from 9 h to 11 h. After collecting

and processing all the data (students’ devices MAC

addresses), the results were following: a) four users

were recognized continuously during the periods of

the lecture on different days (18th, 25th of Septem-

ber, 2nd, 9th of October) (see Table 1). b). All the

users were recognized by the system approximately

every 5 minutes, which is satisfying if we consider

that the beacon was conﬁgured to discover users the

same amount of time. The beacon device could have

been conﬁgured to collect the student attendance af-

ter less amount of time. However, in our case, that

was not necessary since we found convenient that a

ﬁve minute period was more than enough. We have to

take into account that student entered the classroom

pretty much around the same time. On the other hand,

they left the class at different times (September 18 of

Automatic Attendance Control System based on BLE Technology

293

at 10:50 h, September 25 at 11 h, October 2 at 10:41

h, and on the 9th of October at 10:29 h) (see Table

1). Furthermore, two users that were not part of the

class were discovered by the system on October 9th.

This behavior is probably due to the signal overlap-

ping what we explained before in the paper. Also,

the students do not meet the class period requirements

since they were only recognized once in two hour pe-

riod, and we can consider them as not valid for the

speciﬁc lecture attendance list.

Table 1: Recognized users in the classroom.

Classroom 218

08:57 -

10:50

09:01 -

11:00

09:01 -

10:41

09:03 -

10:29

User 18/09 25/09 02/10 09/10

Hanna 25 29 20 16

David 24 27 21 19

Marion 26 31 25 19

Anna 28 30 21 17

Jane 1

Luke 1

Average 5 4 6 7

We have to mention that no inﬂuence on users was

done, regarding their smart devices and activating the

BLE protocol, whatsoever. Also, for the sake of user

privacy, we had to change users MAC addresses for

randomly generated names.

5 CONCLUSION

Even though beacon devices are becoming more pop-

ular and it is evident that lately are included in many

research and business projects, their potential is un-

derutilized. That said, the majority of processing is

handled by the application, which resides on a user’s

smartphone or some other compatible device. How-

ever, letting the software part to do all the process-

ing can become application saturated, and if we take

into account that today’s users have many apps in-

stalled on their device, the problem is evident. In

this work, we presented the solution where a part of

the processing job is shifted and assigned to the bea-

con device. Such method can help us to overcome

the early mentioned problem of possible application

over-processing. It does that by recognizing and serv-

ing simultaneously many user proﬁles and can have a

signiﬁcant potential in any message dissemination or

collecting data environment.

This method of including beacon device into more

active processing and system involvement according

to particular inputs can be used in many areas. It

makes it even easier since Raspberry Pi accepts broad

types of sensors for different requirements and it

makes a perfect device for any beacon solution. That

being said, we give an indicative list of applications

where this solution can ﬁnd a use:

• Space analytics, where users are monitored

through their movements while using building as-

sets. Such analytics are especially useful for

building managers and their future planning of

building expansion and better management of the

building environment.

• Hospital system and indoor localization, where

the patients are divided into different groups and

the hospital personnel are in charge of its patient

group. This way the doctor when in beacon prox-

imity can receive an instant update and the loca-

tion of his patients without any extra processing.

• Any extending of cloud computing to the edge of

an enterprise’s network, better known as Fog com-

puting, where the operation of computing, net-

working services, and storage is facilitated be-

tween end devices and cloud computing data cen-

ters.

In this work, we offer preliminary results, and for

the future work, we plan to implement the system

through all Faculty areas and along with obtaining

student attendance through more extensive tests. We

also expect to pair the dynamic beacon solution with

the corresponding application. The application would

be installed on users’ phones and interact with stu-

dents on a daily basis by sending them interest-based

information. The additional part would be collecting

student movement data that later would be processed

for better space analytics.

ACKNOWLEDGEMENTS

This work is funded by: the European Regional De-

velopment Fund (ERDF) and the Galician Regional

Government under agreement for funding the At-

lantic Research Center for Information and Com-

munication Technologies (AtlantTIC); the Spanish

Ministry of Economy and Competitiveness under the

National Science Program (TEC2014-54335-C4-3-

R and TEC2017-84197-C4-2-R); and the European

Commission under the Erasmus Mundus Green-Tech-

WB project (Smart and Green technologies for inno-

vative and sustainCable societies in Western Balkans;

551984-EM-1-2014-1-ES-ERA MUNDUS-EMA21)

WINSYS 2018 - International Conference on Wireless Networks and Mobile Systems

294

REFERENCES

Apoorv, R. and Mathur, P. (2016). Smart attendance man-

agement using Bluetooth Low Energy and Android. In

2016 IEEE Region 10 Conference (TENCON), pages

1048–1052.

Boric, M., Fernandez Vilas, A., and P. Diaz Redondo, R.

(2018). Analysing the potential of ble to support dy-

namic broadcasting scenarios. International Journal

of Ad hoc, sensor and Ubiquitous Computing (IJA-

SUC).

Komai, K., Fujimoto, M., Arakawa, Y., Suwa, H.,

Kashimoto, Y., and Yasumoto, K. (2016). Elderly

person monitoring in day care center using Bluetooth

Low Energy. In 2016 10th International Symposium

on Medical Information and Communication Technol-

ogy (ISMICT), pages 1–5.

M. Boric, R. P. Diaz Redondo, and A. Fernandez Vilas

(2018). Dynamic Content Distribution over BLE

iBeacon Technology: Implementation Challenges.

Thessaloniki. IEEE Xplore.

Martin, P., Ho, B.-J., Grupen, N., Munoz, S., and Srivas-

tava, M. (2014). An ibeacon primer for indoor lo-

calization: demo abstract. In Proceedings of the 1st

ACM Conference on Embedded Systems for Energy-

Efﬁcient Buildings, pages 190–191. ACM.

Newman, N. (2014). Apple iBeacon technology brieﬁng. J

Direct Data Digit Mark Pract, 15(3):222–225.

Putra, G. D., Pratama, A. R., Lazovik, A., and Aiello, M.

(2017). Comparison of energy consumption in Wi-Fi

and bluetooth communication in a Smart Building. In

2017 IEEE 7th Annual Computing and Communica-

tion Workshop and Conference (CCWC), pages 1–6.

S. Barapatre, A., A. Shelake, V., Y. Pawar, G., and D.R., A.

(2017). Smart College System using IoT BLE Bea-

cons. Internation Jurnal of Advanced Research in

Computer and Communication Engineering, 6. DOI:

10.17148/IJARCCE.2017.6485.

Saraswat, G. and Garg, V. (2016). Beacon controlled cam-

pus surveillance. In 2016 International Conference on

Advances in Computing, Communications and Infor-

matics (ICACCI), pages 2582–2586.

Shende, P., Mehendarge, S., Chougule, S., Kulkarni, P., and

Hatwar, U. (2017). Innovative ideas to improve shop-

ping mall experience over E-commerce websites us-

ing beacon technology and data mining algorithms. In

2017 International Conference on Circuit ,Power and

Computing Technologies (ICCPCT), pages 1–5.

Srinivasan, V. S., Saleem, Kumar, T., and Yasarapu, D. K.

(2016). Raspberry Pi and iBeacons as environmental

data monitors and the potential applications in a grow-

ing BigData ecosystem. In 2016 IEEE International

Conference on Recent Trends in Electronics, Infor-

mation Communication Technology (RTEICT), pages

961–965.

Townsend, K., Cuf, C., and Davidson, R. (2014). Getting

started with Bluetooth low energy: Tools and tech-

niques for low-power networking. ” O’Reilly Media,

Inc.”.

Trifunovic, S., Distl, B., Schatzmann, D., and Legendre, F.

(2011). WiFi-Opp: Ad-hoc-less Opportunistic Net-

working. In Proceedings of the 6th ACM Workshop

on Challenged Networks, CHANTS ’11, pages 37–42,

New York, NY, USA. ACM.

Automatic Attendance Control System based on BLE Technology

295