Murjitama, F.L., et.al.

: Smart Door Lock Using Face Recognition Access 199

Based on Internet of Things (IoT)

Smart Door Lock Using Face Recognition Access

Based on Internet Of Things (IoT)

Farrel Laogi Murjitama1*, Hafidz Nur Raihan2, Rangga Prasetya Adiwijaya3, Desi Fitriani Ramadan4,
Bagas Imanuel Pasaribu5, Bintang A. Silalahi6, Nada Nadiefah Tasman7, Syafira Audri Dwijayanti8, Ummu
Putri Salsabila Panjaitan9, Yudhi S. Purwanto10
1,2,4,5,6,8,10
Informatics Engineering, Faculty of Telematics Energy, Institut Teknologi PLN, Jakarta Barat, DKI
Jakarta, Indonesia
3,7
Electrical Engineering, Faculty of Electricity and Renewable Energy, Institut Teknologi PLN, Jakarta Barat, DKI
Jakarta, Indonesia
9
Information System, Faculty of Telematics Energy, Institut Teknologi PLN, Jakarta Barat, DKI Jakarta, Indonesia

E-mail: 1*[email protected]

(Received: 4 May 2024, revised: 29 May 2024, accepted: 3 Jun 2024)

Abstract

Security is one of the basic things that humans need. In relation to a house or room, the focus of security is on the door lock.
Various types of door locks have been created, ranging from conventional ones with physical keys, to modern types. This modern
type of door lock is also made with various ways to open it. Some use a series of codes (keypad), some use card sensors,
fingerprint sensors, to the use of face recognition technology. Several door lock technologies with face recognition have also
been created, but they are still expensive. The other problem is that those devices are not equipped with some fail-safe
mechanisms, in case there are troubles with the device. This smart door lock is made using face recognition technology based
on the Internet of Things. This lock is equipped with an ESP32cam camera integrated in the ESP8266MOD module that can
recognize faces that have been registered in the database on the website. In addition, the door is also equipped with a push button
to open the door from the inside, and a button as a backup if there is a malfunction of the face recognition feature. The device
test indicates no apparent issues and operates smoothly. The accuracy test for the camera yields positive outcomes, reaching up
to 100% in normal lighting conditions, and dropping to around 60-80% in blur condition. Accuracy is further compromised,
potentially dropping in dim light that the images are only reached 40-60% for clear images and 20% in blurry images.

Keywords: Smart Door Lock, ESP32cam, ESP8266MOD, Push Button, Keypad

I. INTRODUCTION a condition for one of the prerequisites in a dynamic society

for the implementation of the national development process.
Information technology that supports the industrial
Security and order is very important in the continuity of social
revolution 4.0, is fast, efficient and reliable. Based on the
activities, because if it is not controlled it will result in
observations of experts, the rapid progress of information
conditions that are not conducive to the community itself and
technology at this time is very fast and almost complete in all
even the outside community. Lonely places, and lack of
circles and all fields. The use of information systems to support
security can be a source of criminal acts or actions that are
organizational performance is increasingly necessary because
contrary to the ethics and norms of society, such as: theft,
information is an important resource in modern management.
robbery, robbery, even persecution. In this case, it will
Driven by the sophistication of information technology now,
encourage the development of security aspects in society in
this helps the development of increasingly sophisticated
order to prevent noise or acts of bad behavior in the
information systems such as websites or mobile [1].
surrounding environment [3].
In the current era, almost all fields have used information
The importance of security has become a common concern
technology, for instance, in the field of public services. Public
for everyone, given many cases of rampant theft. The National
security and order is the reason in the National Police Chief
Police cracked down on 58,208 people who were reported for
Regulation No. 23 of 2010 [2], which states that Kamtibmas is
theft with aggravation. The motives varied from stealing in

DOI: 10.34148/teknika.v13i2.816 TEKNIKA, Volume 13(2), Juli 2024, pp. 199-203

ISSN 2549-8037, EISSN 2549-8045
200 Murjitama, F.L., et.al.: Smart Door Lock Using Face Recognition Access
Based on Internet of Things (IoT)

empty houses to pick-pocketing. The reported perpetrators make keyless door lock technology more optimal and more
were sentenced to 12 years in prison. The data was obtained economical than other locking methods, so that door security
from the e-MP Robinopsnal Bareskrim Polri [4] for the period can be increased and flexible but with more economical
January 2019 to February 14, 2023. The number of reported devices. Based on the above analysis, the team made a Smart
law-breakers tends to increase from year to year. In 2019, a Door Lock device using IoT-based Face Recognition Access.
total of 7,316 people were reported in connection with theft
crimes. This number increased until 2021. However, in 2022,
the number reported decreased to 15,441 people. Meanwhile, II. RESEARCH METHOD
from January to February 14, 2023, the number was 1,698
The research method used is R&D (Research and
people. This condition then will create an insecure feeling both
Development) with a prototyping development model. The
in personal and society level, and finally will lead to several
stages and activities carried out are the needs analysis stage,
changes in society [5].
design and design of the Smart Door Lock system, making
With so many crimes, we should have a security system
system prototypes, testing until the validation stage. This
that can suppress worry and anxiety, to be able to freely leave
research uses the Research and Development type of research
the house with peace of mind. Human-based security systems
because it is considered in accordance with the objectives of
such as housing security and siskamling may help, but we need
the study, namely to produce and know the results of testing
a tool that is very reliable in doing its job [6].
on the development of IoT-based Smart Door Lock using the
Conventional security that we often encounter to lock the
ESP8266MOD microcontroller (Figure 1).
house includes locking the door leaf and if the user wants to
add some extras to increase security the user can use a padlock
with an additional key [7]. This leads to a lack of efficiency in
locking the door, we can carry the house key but when we lose
the additional lock key then we will struggle in opening the
door. Moreover, the longer the key is use the less accurate the
function, if we buy a cheap door lock, the material used is
usually a mixture that is less qualified for long-term use, often
examples of problems that arise are broken keys when trying
to open the door. This means that to overcome this we need to
also carry a spare key even though we also carry the main key,
because if we put the spare key in the house then we need to
enter the house in a locked state, then the solution is actually
not very effective to solve possible problems that arise in the
future [8].
Figure 1. Research’s Flow Diagram
There have been many developments in locking systems
both mechanically and those that have been integrated with
This activity begins with searching for literature from
more modern technology. Call it keyless technology where the
books, journals, and articles in print and online media. Looking
locking system no longer uses a physical key but is integrated
for the theory of smart door lock application using
with a system. Keyless systems have many types such as finger
ESP8266MOD to open the door automatically. As well as
print-based locking systems [9], [10], key passwords [11], and
conducting literature studies on the calculation of electrical
special cards [12]. All of these strive for guaranteed security as
loads that will be used, especially DC electrical power
well as flexibility. However, it should be realized that
management systems. Furthermore, calculations and system
technology-integrated locking has a price that is not cheap, this
becomes market segmented with differences in economic modeling are carried out to obtain a Smart door lock prototype.
The first calculation aims to determine how much electrical
value, so people still stick with conventional locking systems.
power consumption. The second calculation is done to find out
Face recognition doorlock was also made and researched,
how much DC voltage is needed.
such as by [13], [14], and [15], but those researches do not
provide a safe plan if in their implementations, those systems
experience failures. Some failures that might occur, such as
failed circuit, drained battery, or jam in its locking mechanism.
As a modern locking solution, the team designed a door
locking system that will combine Internet of Things (IoT) and
Artificial Intelligent (AI) technology. The operation concept of
this design is to use ESP32cam for an AI-based facial Figure 2. Smart Door Lock Scheme
recognition system that can be reached on a web application
(websoftware) as an integrated security access rights regulator. In designing a smart door lock using IoT-based face
This system is also equipped with an ESP8266MOD, based on recognition, it consists of various system reviews as shown in
IoT that can be connected wirelessly as a concept of data the Figure 2 above which includes: 1) Face recognition is a
communication between devices. This idea is expected to system design from a camera device that main feature is to scan

TEKNIKA, Volume 13(2), Juli 2024, pp. 199-203 DOI: 10.34148/teknika.v13i2.816

ISSN 2549-8037, EISSN 2549-8045
Murjitama, F.L., et.al.: Smart Door Lock Using Face Recognition Access 201
Based on Internet of Things (IoT)

faces to match database faces and regulate access (Figure 3 and The process of designing the tool circuit is carried out after
4), 2) The fail-safe button is an alternative security design to the calculation and modeling of the system and 50% of the tool
open the door without using the IoT. For example, if some design has been completed. An overview of the technology to
electrical problems happened, or when the person is not be developed is below. The design of the tool circuit is divided
familiar to the system. into 3 main parts, namely: ESP programming, power-signal
distribution system, and mockup design (miniature simple
door) as shown in Figure 5 and 6.
The testing phase is carried out to measure the success of
the smart door lock as a modern door lock, which is minimalist,
economical, portable, reliable, and efficient. Testing was
Figure 3. Illustration of the ESP Device Programming Process carried out including to find out: 1) The level of success, and
ease of access controlled by the Smart Door Lock electronic
system, to open the door lock (solenoid lock, and ESP control
system), and 2) Testing backup access as a backup-plan if a
problem occurs when the door is locked, where the smart door
lock must continue to function in various conditions, for the
Figure 4. Illustration of Face Registration Process in Face sake of room security.
Recognition Database Report writing is carried out after all stages are completed
so that the results obtained from the design of the tool can be
Furthermore, 3) Door lock is a door lock system that uses explained in detail and specifically in accordance with the data
a solenoid device. This system is equipped with an obtained. The results of the report will be submitted to clients
electromagnet as a Solenoid lock driver, 4) Smart Door Lock and examiners for consideration.
is a door lock design using IoT access, 5) Electricity
management is a design to regulate electricity flow so that it
can meet the electricity needs of the load, 6) Software is an III. RESULTS AND DISCUSSIONS
SDL system design, and 7) Control is the design of software The tool development process is divided into 2 parts,
integration and Solenoid lock. namely the part of face recognition software (Figure 7) and the
part of making and installing the tool on the door (Figure 8).
In making face recognition software, the interface design is
made in accordance with the existing design, so that the
resulting display becomes as below:

Figure 7. Face Recording Display in the Face Recognition

Application
Figure 5. Inside Installation Sketch of the Device

Figure 6. Outside Installation Sketch of the Device Figure 8. Circuit Sketch Result

DOI: 10.34148/teknika.v13i2.816 TEKNIKA, Volume 13(2), Juli 2024, pp. 199-203

ISSN 2549-8037, EISSN 2549-8045
202 Murjitama, F.L., et.al.: Smart Door Lock Using Face Recognition Access
Based on Internet of Things (IoT)

Tests were conducted in several conditions, including IV. CONCLUSIONS

lighting (sufficient and low light), front face side, and clipped
An automatic door locking device has been made at a
objects. Tests were conducted on 3 objects with 5 tests each
relatively low cost even though it has technology that is in
with different but similar objects as shown in Table 1-3 below.
accordance with current technological developments. Both the
device and the facial recognition application function can
Table 1. Testing Results Data of Personnel A
function properly as long as the light conditions are good
Object
enough with a success rate of 100% in 3 different people with
No CONDITION Personnel A 5 trials. The keypad function as a backup door opener also
T F Acc works well, indicated by 100% success on 10 trials.
In a well-lit room Disadvantages exist in the quality of the camera which will
1 a Front side 5 0 100% decrease its performance when light conditions decrease, and
b Blur object 3 2 60% the condition of the camera is not focused (blur). Making a
smart door lock with better camera quality will increase the
In a low-lit room performance of this tool, so it will be able to make it more
2 a Front side 2 3 40% perfect.
b Blur object 1 4 20%
Capt.: T = True, F = False, Acc. = Accuracy
ACKNOWLEDGEMENT
Table 2. Testing Results Data of Personnel B
The authors would like to thank the Research and
Object
Community Service Bureau (LPPM) Institut Teknologi PLN
No CONDITION Personnel B
(ITPLN) for providing support that made this research
T F Acc possible.
In a well-lit room
1 a Front side 5 0 100%
b Blur object 4 1 80% REFERENCES

In a low-lit room [1] D. F. Sumajouw, M. E. I. Najoan, and S. R. U. A.

2 a Front side 3 2 60% Sompie, “Perancangan Sistem Keamanan Rumah
b Blur object 1 4 20% Tinggal Terkendali Jarak Jauh,” J. Tek. Elektro dan
Komput., vol. 4, no. 3, pp. 44–53, 2015.
Capt.: T = True, F = False, Acc. = Accuracy
[2] B. H. Danuri, “Peraturan Kapolri No. 23 Tahun 2010.”
2010.
Table 3. Testing Results Data of Personnel C
[3] M. Saleh and M. Haryanti, “Rancang Bangun Sistem
Object
Keamanan Rumah Menggunakan Relay,” J. Teknol.
No CONDITION Personnel C
Elektro, Univ. Mercu Buana, vol. 8, no. 2, pp. 87–94,
T F Acc 2017, [Online]. Available:
In a well-lit room https://media.neliti.com/media/publications/141935-
1 a Front side 5 0 100% ID-perancangan-simulasi-sistem-pemantauan-p.pdf.
b Blur object 3 2 60% [4] “e-MP Robinopsnal Bareskrim Polri.pdf.” .
[5] H. Ocktilia, Y. S. Ismudiyati, and B. Rusyidi, “Social
In a low-lit room Transformation of Rural Communities Towards
2 a Front side 2 3 40% Independent Prosperous Village in Indonesia,” … Conf.
b Blur object 1 4 20% Soc. …, no. March, 2020.
Capt.: T = True, F = False, Acc. = Accuracy [6] M. S. Sungkar, T. Elektronika, P. Harapan, and B.
Tegal, “SISTEM KEAMANAN RUMAH BERBASIS
The test results of the device on several people whose faces INTERNET OF THINGS,” vol. 9, no. 2, pp. 1–3, 2020.
have been recorded show that the system and its app have [7] M. Derbali, “Toward Secure Door Lock System:
shown their jobs with a 100% success rate in normal room light Development IoT Smart Door Lock Device,” pp. 1–28,
conditions but decreased in the results of recording unfocused 2023, [Online]. Available:
faces by 40% (60% success rate). While in low light https://doi.org/10.22541/au.168055385.54003954/v1.
conditions, the results obtained will further decrease, even on [8] S. Wankhade, S. Gurbhele, D. Apage, and C. Bore,
the front face object only up to 40% success and 20% on the “Smart Door Lock System Using Blockchain Based on
unfocused object. Solana Technology,” Int. J. Eng. Appl. Sci. Technol.,
vol. 8, no. 3, pp. 85–88, 2023, doi:
10.33564/ijeast.2023.v08i03.011.
[9] H. Yadu, S. Purbey, and A. K. Choudhary, “Fingerprint
Door Lock,” Interantional J. Sci. Res. Eng. Manag., vol.

TEKNIKA, Volume 13(2), Juli 2024, pp. 199-203 DOI: 10.34148/teknika.v13i2.816

ISSN 2549-8037, EISSN 2549-8045
Murjitama, F.L., et.al.: Smart Door Lock Using Face Recognition Access 203
Based on Internet of Things (IoT)

06, no. 05, pp. 1–4, 2022, doi: 10.55041/ijsrem13541. Sejal Zade, and Hemlata Dakhore, “Face Recognition
[10] Andi Zulkifli Nusri and Ashar Alimuddin, “Rancang Based Door Lock System,” Int. J. Sci. Res. Sci.
Bangun Security Door Lock Menggunakan Fingerprint Technol., pp. 395–401, 2022, doi:
Berbasis Mikrokontroler,” J. Ilm. Sist. Inf. dan Tek. 10.32628/ijsrst229268.
Inform., vol. 5, no. 1, pp. 23–28, 2022, doi: [14] S. Qureshi, “Face Recognition (Image Processing)
10.57093/jisti.v5i1.105. based Door Lock using OpenCV, Python and Arduino,”
[11] J. Al-rasyid, M. Nawawi, and C. P. Pratiwi, “Keypad International Journal for Research in Applied Science
Passcode Design Analysis on Smart Lock Door System and Engineering Technology, vol. 8, no. 6. pp. 1208–
IoT Based,” vol. 16, no. 2, 2023. 1214, 2020, doi: 10.22214/ijraset.2020.6197.
[12] D. R. Azzizi and Rino, “Rancang Bangun Sistem [15] S. Abdullah, “Development of Face Recognition Based
Keamanan Door Lock Menggunakan E-Ktp Berbasis Smart Door Lock,” International Journal of Advances
Arduino Wemos Esp8266,” Algor, vol. 4, no. 1, pp. 21– in Scientific Research and Engineering, vol. 07, no. 07.
27, 2022, doi: 10.31253/algor.v4i1.1742. pp. 01–05, 2021, doi: 10.31695/ijasre.2021.34036.
[13] Achal Meshram, Chetna Kumbhare, Snehal Chaudhari,

DOI: 10.34148/teknika.v13i2.816 TEKNIKA, Volume 13(2), Juli 2024, pp. 199-203

ISSN 2549-8037, EISSN 2549-8045