DEV Community

INSU SHIN
INSU SHIN

Posted on

Implementation of a Smart Sensor for Proactive Fire Prediction

  • Research Background Recently, there has been a frequent occurrence of large-scale fires in traditional markets and factories caused by negligent management, equipment defects, and aging infrastructure. In particular, instantaneous sparks and over currents resulting from deteriorated wires or poor contact are the fundamental causes of such major fires.

Currently, conventional circuit breakers operate on a reactive mechanism, cutting off the entire power supply only after an over current or short circuit exceeds a specific threshold. The clear limitation of these traditional systems is their inability to proactively detect minute sparks and leakage currents that trigger fires. Furthermore, when a problem does occur, they cannot pinpoint the exact location of the fault, leading to delayed responses.

To overcome these limitations, we determined that it is necessary to introduce a smart sensor capable of anticipating and warning of fire risks by analyzing magnetic field variations and over currents around wires before a fire breaks out. This research was conducted to directly implement this system and verify its reliability, with the goal of enhancing safety and ensuring accessible use for everyone.

  • Research Objectives The ultimate objective of this study is to improve upon the limitations of the reactive cutoff mechanism in existing circuit breakers by implementing and verifying a clamp-type sensor that proactively detects internal wire conditions and abnormal signals prior to a fire.

To validate the proposed circuit, this study proceeds with circuit design and simulation using LT spice, followed by the physical fabrication of the circuit to confirm its ability to detect magnetic field variations and sustained abnormal signals.
Additionally through this cross-validation process, the study examines the following practical considerations "Is it applicable in real-world scenarios?", "Can abnormal signals be easily identified?", and "Is it convenient for anyone to use?“

  • Components Used and Circuit Configuration
    In this study, not all the components initially proposed in the project plan could be utilized; instead, the circuit was constructed using alternative components.

  • Exclusion of Hall Sensors in Favor of Coils
    Originally, we planned to use both a coil and a Hall sensor to detect magnetic fields, but only a coil was utilized in this study.
    The reason lies in the operational characteristics of the Hall sensors we intended to use.

While they are capable of detecting magnetic field variations, they are digital components that output data exclusively as 1 and 0.
Because our system required continuous analog signals rather than discrete binary outputs, the Hall sensor was ultimately excluded.

Instead, considering the final form factor of the sensor, a custom coil was fabricated and integrated into the circuit by carefully calculating its cross-sectional area, number of turns, and total length.

  • Implementation of a Clipping Circuit A clipping circuit was configured using rectifier diodes and Zener diodes. This setup allows the magnetic field variations and internal AC signals from the wire, picked up by the coil, to pass into the circuit while preventing damage caused by intense sparks or high-voltage transients. The primary purpose of the clipping circuit is circuit protection, achieved by cutting off (clipping) any signal portion that exceeds a predefined reference voltage.

Originally, the plan was to implement a full-wave rectifier circuit. However, this was modified after clearly distinguishing the functional differences between the two topologies. A full-wave rectifier serves to convert highly irregular AC signals into DC signals.
However, if an extremely severe spark or arc inside the wire passes through a full-wave rectifier, the exceptionally large AC signal is simply converted into a massive DC signal and passed along, which can severely compromise the stability of the subsequent circuitry.

In other words, because a rectifier fails to protect the circuit from voltage spikes, we determined that a clipping circuit—which protects the system by capping any signal exceeding the threshold at a predetermined value—was far more suitable for our research objectives.

Furthermore, while TVS (Transient Voltage Suppressor) diodes are typically available for voltage suppression in such configurations, the circuit was constructed using rectifier diodes and Zener diodes due to cost constraints.

  • Replacement of Synaptic Transistors with a Leaky RC Integrator Circuit Initially, we planned to utilize synaptic transistors to process the signals passing through the coil and clipping circuit, enabling the system to distinguish between normal and abnormal signals. However, because fabricating synaptic transistors requires processes beyond the capabilities of an undergraduate-level laboratory, we replaced them with a leaky RC integrator circuit that yields an equivalent operational effect.

This circuit utilizes the charging and discharging characteristics of a resistor and capacitor to achieve a "cumulative detection effect."
It internally accumulates and "remembers" sustained abnormal signals, which eventually alters the output voltage. Additionally, it features an "instantaneous over voltage detection" capability. Even if a spark enters the circuit for a very brief duration, if its amplitude is sufficiently large, it generates a value equivalent to a continuously sustained signal, enabling immediate detection.

A critical question raised during this design phase was "If the system detects anomalies by accumulating signals, wouldn't normal signals also accumulate over time and trigger a false detection?"
The answer lies in the charging and discharging time constants of the capacitor. While normal signals also pass through the circuit as AC signals, their voltage amplitude is extremely small. Even as the capacitor charges, the rate of discharge through the resistor is significantly faster than the rate of charge. Consequently, the accumulated voltage never exceeds the predefined threshold, effectively preventing false alarms.

Therefore, although the circuit architecture differs from the originally planned synaptic transistors, it successfully replicates the desired detection mechanisms while preventing malfunctions, thereby ensuring stable signal input and output processing.


  • Output Stage and Signal Amplification The final component of the system is the output section, which provides visual confirmation if an abnormal signal is detected from the preceding stages.

Visual Notification via LED
An LED was utilized for this purpose.
When an anomaly is detected, the LED illuminates, enabling immediate visual identification of the abnormal signal and allowing for a real-time assessment of the internal condition of the wire. Initially, the design included both an LED and a buzzer for simultaneous visual and auditory alerts.

However, due to procurement issues with transistor components, the buzzer was excluded, and the final circuit was configured solely with the LED.

Amplification Limitations and Future Improvements
Driving a buzzer requires a higher amplification factor, which would typically be achieved by utilizing BJT or MOSFET components.
Because of the aforementioned supply issues, an Operational Amplifier (OP-AMP) was used for signal amplification instead.
While an OP-AMP is capable of amplification similar to discrete transistors, its output current drive in this configuration is only sufficient to power an LED.
It exhibits a distinct limitation in driving capability for higher-power loads, such as buzzers or vibration sensors.

Nevertheless, this limitation is not considered a critical flaw in the overall design. If this project is selected for formal subsequent research or if budget constraints are resolved, the issue can be easily rectified by simply substituting the relevant amplifying components within the existing circuit architecture.



  • Verification of Sensor Reliability via Simulated Fire Environments To verify the stability and normal operation of the custom-designed sensor, two methods of cross-validation were conducted.

The first method utilized a piezoelectric element to simulate the sparks and arcs that typically occur immediately before a fire breaks out.
A piezoelectric element, extracted from a common lighter, was applied to the front end of the coil to repeatedly inject instantaneous high voltages and rapid magnetic field variations into the circuit.

We determined that this component could closely replicate the instantaneous high-voltage sparks generated in actual fire scenarios.
Direct measurements revealed that one or two sparks or arcs did not trigger the LED.
Instead, it required a minimum of five to six consecutive sparks for the LED to illuminate.

This result perfectly aligned with our theoretical expectations.
It confirmed that the clipping circuit successfully protected the system by suppressing excessively high voltages, while the leaky RC integrator circuit accumulated ("remembered") the sustained abnormal signals, eventually reaching the threshold voltage to drive the LED.




The second validation utilized a function generator to determine whether a "single, massive spark and rapid magnetic field change"—as opposed to a continuous signal—could also successfully trigger the LED.
Because generating such an extreme single event is difficult using only a piezoelectric element, a function generator was employed to verify the outcome.

A pulse wave with a high frequency and a high peak-to-peak voltage was generated and applied to the circuit.
The clipping diodes successfully protected the circuit by suppressing the high-frequency transients. By utilizing rapid ON/OFF switching to inject a single, powerful impulse, we confirmed that the LED illuminated as intended.


  • Final Circuit Design and Sensor Form Factor The aforementioned circuit blocks were integrated into a single cohesive system, and the physical sensor was designed in a clamp-type form factor to ensure ease of use for anyone.

The circuit schematic was initially verified using LTspice to evaluate operational feasibility and calculate component costs.
In the second phase, the prototype was constructed on a large breadboard to physically confirm its normal operation.
Finally, to achieve miniaturization, the circuit layout was optimized and transferred to a small breadboard. This step successfully minimized the overall volume while maintaining full functionality.




The final sensor prototype features a non-contact, clamp-type design.
This configuration eliminates the need to strip wire insulation or power down the system for inspection simply clamping the sensor onto a suspected wire allows for immediate assessment of internal electrical conditions.
Additionally, the LED indicator is positioned in the most visible location to ensure that hazard alerts can be recognized instantly.

  • Limitations and Future Expectations (Conclusion) In this study, we designed and implemented a "non-contact, clamp-type fire prevention sensor" capable of detecting internal wire anomalies. This system proactively prevents fires, thereby overcoming the fundamental limitations of the reactive cutoff mechanisms found in conventional circuit breakers. Through a four-stage circuit topology—comprising a coil, a clipper, a leaky RC integrator, and a comparator—the system successfully detected both the gradual accumulation of minute abnormal signals and instantaneous high-voltage sparks.

In particular, through cross-validation experiments utilizing a function generator and a piezoelectric element to simulate extreme arcs and sparks, we proved that the circuit remains physically protected while outputting accurate alerts, even under conditions mirroring harsh field environments.

This sensor can be easily attached to aging wiring without damaging the insulation or requiring power outages. It presents an economical and highly practical preventive solution with broad applicability, ranging from everyday residential use to fire-prone environments like traditional markets.

  • Despite its successes, there is room for further enhancement. A primary future objective is to introduce power transistors, such as BJT or MOSFET, to improve the output current drive.

This upgrade would enable the integration of high-power buzzers or vibration motors, evolving the device into a multi-alarm sensor that can effectively alert individuals with visual or auditory impairments, as well as workers in high-noise industrial environments.

Furthermore, by incorporating wireless communication technology, the system could be upgraded into a next-generation safety network capable of transmitting real-time hazard warnings directly to an administrator's smartphone.
We expect that these advancements will significantly contribute to reducing fire incidents across society.

Top comments (0)