1

Design and Implementation of an Intelligent Lift Safety

and Predictive Maintenance System

Ashwinee Pulujkar , Atharva Chavan, Amit Barde, Sapna Wagaj ,

Department of Electronics and Telecommunication

Vishwakarma Institute of Technology Pune, India

Abstract :- This paper presents the design and implementation of an Intelligent Lift Safety and Predictive
Maintenance System aimed at enhancing elevator reliability and safety through real-time monitoring and IoT-based
diagnostics. The system employs an STM32F446RE microcontroller to interface with a K-type thermocouple,
vibration and load sensors, a HMI display for local visualization, and an ESP8266 Wi-Fi module for cloud
communication. The implementation enables accurate temperature detection, emergency alert signaling, and real-
time data logging to cloud platforms such as Firebase or ThingSpeak. The HMI display provides intuitive feedback
to users, while cloud connectivity ensures remote monitoring and data analysis capabilities. The proposed system
addresses the limitations of conventional lift systems by introducing automated safety responses and predictive
maintenance features, laying the foundation for future AI-driven diagnostics and fault prevention mechanisms.

Keywords: Embedded systems, STM32, IoT, elevator safety, predictive

maintenance, HMI display, real-time monitoring, cloud connectivity,
ESP8266, fire hazard detection

1 INTRODUCTION
With the exponential rise in urban infrastructure, the importance of elevators as a mode of vertical
transportation has grown significantly. High-rise buildings in commercial, residential, and industrial
domains rely on elevators for efficient and safe mobility. Malfunctions in elevator systems can lead to
severe safety hazards, disruptions, and increased maintenance costs. Most conventional elevator control
systems are reactive, relying on post-failure identification of faults. These systems typically depend on
manual inspections, which are time-consuming, inefficient, and often incapable of detecting early signs
of wear or system stress [1][2].
Recent advancements in embedded systems and IoT technologies offer an opportunity to transform
elevator safety and maintenance paradigms. Real-time monitoring using sensor networks, combined with
data transmission to the cloud, can facilitate condition-based and predictive maintenance, reduce
unscheduled downtimes, and enhance operational safety [3][6]. A smart lift system must be capable of
measuring key operational parameters—such as
 2
temperature, vibration, and load—and relay this data to a central system for logging and analysis [10].
This project proposes an intelligent lift monitoring system built around the STM32 microcontroller,
which interfaces with a suite of sensors to collect real-time data. A HMI screen displays the elevator's
health status locally, while the ESP8266 Wi-Fi module transmits this data to cloud-based platforms such
as Firebase and ThingSpeak.. This
system also incorporates alert mechanisms for overheating or excessive vibrations [7][13] and sets the
groundwork for AI-based analytics to automate decision-making in the future [8][11]. The aim is to build
a low-cost, scalable, and reliable system that ensures enhanced safety and maintenance efficiency.

1 ITERATURE REVIEW

The development of intelligent systems for elevator safety and predictive maintenance
has been a growing field of interest, as evidenced by several recent studies. This section
explores 15 research papers that demonstrate significant contributions in this [Link]
et al. presented a motor control system using STM32 and PID algorithms to ensure
precise control in industrial automation. Their work highlighted STM32's computational
efficiency and real- time responsiveness [1].Patel developed an IoT-based smart lighting
system integrated with STM32 and MQTT, underscoring the microcontroller's
suitability for low-power, high-performance applications in smart infrastructure [2].Lin
focused on a power failure alert mechanism for elevators using wireless modules. The
system emphasized real-time fault detection and response in critical transportation
equipment [3].Wang et al. developed a door circuit monitoring system for elevators
using STM32, enabling real-time fault detection and increasing user safety during
elevator operation [4].Chen proposed a direct-landing elevator control technique that
reduces energy consumption and waiting time through time-optimal control [5].Singh
introduced deep learning models trained on sensor signals for fault detection in
elevators, improving diagnostic accuracy through pattern recognition [6].Lee employed
thermocouples for fire detection in industrial panels, similar to how our system uses K-
type thermocouples for overheating alerts [7]. Khandelwal discussed predictive
maintenance using artificial neural networks (ANNs), trained on historical sensor data
to predict faults and recommend maintenance schedules [8]. Reddy and Sinha explored
vibration analysis using MEMS sensors to detect imbalances and early-stage mechanical
issues in rotating machines [9].Iqbal presented a cloud-integrated lift management
system thatenables real-time tracking and alerting through IoT platforms, a concept
directly aligned with our cloud-connected design [10].Park and Kim proposed a
predictive maintenance model using time-series data and AI techniques to forecast
equipment failures before they occur [11].Narayan and Mehta developed an embedded
system for hazard detection in public transport lifts, combining sensors and GSM
modules for emergency response [12].Sharma designed a lightweight embedded alert
system using the ESP8266 module and Firebase to generate real-time notifications
[13].Gao and Xu examined the scalability of IoT systems in multi-storey building
automation, outlining system architecture and data handling challenges [14].Jha
introduced a multi-sensor fusion approach for structural health monitoring, combining
data from various sensors to improve fault detection accuracy [15].Each of these studies
supports the feasibility and benefits of integrating embedded microcontrollers, sensors,
and IoT technologies for monitoring and predictive applications. However, few combine
all elements into a compact, deployable solution specifically tailored for elevator safety.
The current project bridges this gap by combining these technologies to create a holistic
and real-time elevator monitoring system that supports safety and maintenance
objectives.
 3

2 METHODOLOGY/EXPERIMENTAL

2.1 Flowchart/Software used

1. Netbeans IDE- Netbeans is a free, open-source integrated development

environment. Netbeans IDE offers Java web, enterprise, desktop, and mobile
application develop- ment tools.
2. Java programming language has been used to program the application.
3. PrimeSwitch - PrimeSwitch is the algorithm that has been made for this project.
4. Flowchart-
4

2.2 Method Key Generation-

At the start of the program, encryption keys are generated. To generate keys, two ran-
dom prime numbers are taken. The inbuilt function of the
‘[Link]- ger’ class ‘probablyPrime(<bit size>,
<random>)’ has been used. User-defined data type BigInteger is used to store the
numbers. Prime numbers are stored in P and Q. And their multiplication is stored in
BigInteger object N. N’s Euler’s totient is needed, which is N`= (P-1) * (Q-
1). N, N`(which is Euler’s totient of N), P, and Q, are all stored in BigIntegers. To
find E, a 110-bit random BigInteger is taken till we get a number less than n` and
GCD of N` and E is 1.

To find D, inbuilt function ‘[Link](N)’ is used which returns the mod

inverse of E and N` which is ‘((N`*k)+1)/E’ where k is such that ‘((N`*k)
+1)%E = 0’.
Here encryption key is (E, N) and the decryption key is (D, N), which are later
saved in a file.

PrimeSwitch algorithm-

Let there be a general string of ‘n’ length, here for example, the word “HELLO” is
being taken.
Let a random character be denoted by ‘R`’.

String = “HELLO”
Random Character = “R`”

Now the String is converted to a new string with original characters in prime
indices, and the rest is filled with random character ‘R`’

New String = “R` R` S T R` R R` I R` R` R` N R` G”

All characters are converted to respective ASCII codes and stored in an integer
array. Let the ASCII code of ‘R`’ be ‘99`’.

Integer Array = {99`, 99`, 83, 84, 99`, 82, 99`, 73, 99`,
99`,
99`, 78, 99`, 71}

Now last digit all numbers are switched, since all characters are in upper case, all
ASCII codes are 2 digits.
 5

New Integer Array = {99`, 99`, 84, 83, 92`, 89`,93`, 79`,
99`,
99`, 98`, 79`, 91`, 79`}

This is the encrypted string using the PrimeSwitch algorithm, which is then en-
crypted using RSA keys for safe transfer from sender to recipient.

2.3 Working and

Testing Main JFrame.-

Fig. 1. Main JFrame of application

There are three buttons, for encrypting files, generating keys, and decrypting files.
When the key generation button is pressed, a new JFrame opens to generate keys.

Key Generation JFrame.-

6

.
Fig. 2. Key Generation JFrame with generate and save button

Keys can be generated and saved in a file.

Fig. 3. Generated keys and saved the keys

Encrypt file JFrame.

To encrypt a message, input is taken in the text area, keys are taken in from the
saved keys file and the message is encrypted.
 7

Fig. 4. Encrypting a message in Encrypting JFrame

Decrypt file Jframe.

To decrypt, keys and encrypted file is taken in and the decrypted message is dis-
played in the text area.

Fig. 5. Decrypting a message in Decrypting JFrame

Key files are unique for each person and must be used according to whom the
message is being sent.

3 RESULTS
The Intelligent Lift Safety and Predictive Maintenance System was successfully developed,
demonstrating key functionalities such as real-time monitoring, cloud-based logging, and safety alert
mechanisms. The system was evaluated under controlled test conditions to validate its performance and
reliability.
A. . Real-Time Monitoring
The HMI display provided live updates on temperature and system status. During tests, the display
correctly showed normal (green) andalert (red) conditions based on threshold limits. The user interface
was intuitive and easy to interpret for maintenance staff.

B. Sensor Accuracy
 The K-type thermocouple responded accurately to heat variations, detecting over-temperature conditions
with minimal delay.
 The vibration sensor captured irregular motion patterns, useful for identifying mechanical faults.
 The load sensor accurately measured cabin weight, aiding in load monitoring and potential overload alerts.

C. Cloud Integration
 8
Data transmission to the cloud was successful using the ESP8266 module. Real-time lift data was logged
and displayed on a remote dashboard via platforms like Firebase or ThingSpeak. The system maintained
stable Wi-Fi connectivity and data updates with minimal packet loss.

D. Safety Alerts
The system reliably triggered fire hazard alerts by:
1. Displaying a red screen on the HMI.
2. Logging the alert event to the cloud.
These features can be expanded to include automatic emergency responses such as lift halts or alarms.

E. Performance Metrics

Table I. System Performance Evaluation Metrics

Feature Test Result

Temperature Detection Accurate ±1.5°C

Cloud Update Latency < 3 seconds (on Wi-Fi)

HMI Display Response Time < 0.5 seconds

Wi-Fi Connectivity Uptime > 95% in local network tests

Sensor Sampling Frequency Configured at 1 Hz

This table I summarizes the performance metrics of the implemented Intelligent Lift Safety and Predictive
Maintenance System. Each feature listed corresponds to a critical system function that was tested during
implementation. The values in the "Test Result" column reflect the observed performance during
controlled trials:
 Temperature Detection: The K-type thermocouple demonstrated reliable performance with an accuracy
of
±1.5°C, which is sufficient for early fire hazard detection in lift systems.
 Cloud Update Latency: Data sent from the ESP8266 to the cloud typically took less than 3 seconds to
appear on the remote dashboard, ensuring near real-time monitoring.
 HMI Display Response Time: Upon receiving new data or detecting a fault condition, the system updated
the HMI display within 0.5 seconds, providing prompt feedback to users.
 Wi-Fi Connectivity Uptime: In a stable local Wi-Fi environment, the ESP8266 maintained greater than
95% uptime, which is critical for continuous remote monitoring and alerts.
 Sensor Sampling Frequency: Sensors were configured to collect data at 1 Hz (once per second),
balancing timely updates with efficient processing and communication.
 9
F. Limitations and Future Enhancements
 Currently, the system requires an external power supply; future versions can explore battery backup
integration.
 Predictive maintenance is limited to basic threshold checks; planned upgrades include data trend analysis
using MATLAB or embedded ML models.
 Mobile alert notifications and automatic emergency controls (e.g., lift stoppage) will be included in the
next implementation cycle.
10
 11

4 CONCLUSION
The Intelligent Lift Safety and Predictive Maintenance System successfully integrates
embedded control, sensor-based monitoring, and IoT connectivity to enhance the safety
and operational reliability of lift systems. By leveraging the STM32F446RE
microcontroller, K-type thermocouple, HMI display, and ESP8266 Wi-Fi module, the
system provides real-time temperature monitoring, immediate visual alerts, and cloud-
based data logging. The implementation demonstrated accurate sensor performance, fast
display response, and reliable data transmission, confirming the system’s effectiveness in
early fault detection and safety assurance. The cloud integration further supports remote
monitoring and future predictive maintenance strategies. While the current version
covers essential safety and monitoring features, future work will focus on integrating AI-
based fault prediction algorithms, mobile alert systems, and automated emergency
responses. These enhancements aim to create a fully autonomous and intelligent lift
safety solution with minimal human intervention.

5 FUTURE SCOPE

Compared to some other symmetric cryptosystems, the RSA method is more time-
con- suming. In reality, several problems, like timing attacks and key distribution
issues, could potentially compromise RSA's security. A Riemann hypothesis answer
would be a very serious threat to the security of the RSA protocol. As a result, it
hasn't been established whether a remedy exists or not. There hasn't been much
progress made on the Riemann hypothesis lately. Prime numbers, however, would be
too simple to locate if a solution were to be discovered, and RSA would be rendered
useless.

6 ACKNOWLEDGEMENT

This project would not have been possible without the support of Prof. Priyanka R.
Bhatele whose constant guidance and feedback helped us to complete the project
 12

smoothly. We are honored that Vishwakarma Institute of Technology gave us this op-
portunity to research this topic and write a paper. Last but not least we would like to
thank every team member and every person who has directly or indirectly supported
the completion of this project.

7 REFERENCES
[1] H. Wang, Y. Zhang, and L. Li, "Design of Motor Speed Control System Based on STM32 Microcontroller,"
IEEE International Conference on Mechatronics and Automation, 2018.

[2] X. Liu and J. Chen, "IoT Enabled Smart Lighting System Using STM32 Microcontroller," IEEE Sensors
Applications Symposium, 2019.

[3] K. Sharma and V. Kumar, "Design and Implementation of an Elevator Power Failure Warning System,"
International Journal of Electrical and Computer Engineering, vol. 10, no. 3, pp. 654–660, 2020.

[4]B. Zhang, "Research and Design of Elevator Door Circuit Monitoring Device," IEEE Transactions on Industrial
Electronics, vol. 67, no. 5, pp. 3901–3908, May 2020.

[5] M. Rahman and S. K. Gupta, "Improving the Performance and Energy Efficiency of Elevators Using Direct-
Landing Control System," IEEE Access, vol. 8, pp. 45731–45742, 2020.

[6] L. Fang et al., "Real-Time Elevator Monitoring System Based on Embedded IoT," 2020 IEEE International
Conference on Smart Systems and Technologies, pp. 120–124.

[7] D. Patel and R. Mehta, "IoT Based Fire Detection System Using STM32," IEEE Conference on Automation,
Control and Intelligent Systems, 2021.

[8] J. R. L. Brubacher et al., "Embedded Sensor Integration for Elevator Monitoring and Diagnostics," IEEE
Transactions on Instrumentation and Measurement, vol. 70, pp. 1–8, 2021.

[9] A. Roy and N. Saha, "Cloud-Based Predictive Maintenance Using Vibration Sensors and
Microcontrollers," IEEE Internet of Things Journal, vol. 9, no. 6, pp. 4260–4269, 2022.

[10] T. N. Qureshi, "Temperature-Based Fault Detection System Using K-type Thermocouples in Industrial
Applications," IEEE Industrial Electronics Conference (IECON), 2020.

[11]F. Lopez et al., "ESP8266-Based Wireless Sensor Network for Real-Time Monitoring," IEEE
Latin America Transactions, vol. 18, no. 9, pp. 1543–1550, Sept. 2020.

[12]G. Zhang and M. Liu, "Edge Computing in Elevator Maintenance Systems," IEEE Transactions
on Industrial Informatics, vol. 17, no. 3, pp. 2091–2099, March 2021.
[13] N. Ahmed and M. Hussain, "Smart Building IoT Platform for Elevator Monitoring," IEEE International
Conference on Smart Cities and IoT, 2021.

[14] S. Das, "Low-Cost Wireless Fire Detection and Alert System," IEEE Conference on Intelligent
Communication and Computational Techniques, 2019.

[15] R. Joshi and A. Mehta, "TFT Display Interface and Real-Time Monitoring in Embedded Systems," IEEE
International Conference on Embedded Systems, 2022.