Performance Evaluation and Comparative Analysis of Arduino-Based Sensors for

Real-Time Applications
Atharv Mishra
Department of Computer Science and Engineering, Sagar Institute of Research and Technology,
Bhopal, Madhya Pradesh, India
Email: mishra.atharv467@[Link]

ABSTRACT: The proliferation of Internet of Things (IoT) devices has driven significant demand for reliable, cost-effective, and energy-
efficient sensors capable of operating in real-time environments. Arduino-based platforms, particularly the Arduino Uno, have emerged as
widely adopted microcontroller boards for sensor-integrated systems. This paper presents a comprehensive performance evaluation and
comparative analysis of three major sensor categories interfaced with Arduino Uno: (i) Temperature and Humidity sensors (DHT11 and
DHT22), (ii) Motion and Ultrasonic sensors (HC-SR04 and PIR HC-SR501), and (iii) Gas and Air Quality sensors (MQ-2, MQ-7, and MQ-
135). Key performance parameters including measurement accuracy, response time, sampling rate, power consumption, detection range, and
cost-effectiveness are systematically examined and quantified. Results show that DHT22 achieves ±0.5°C temperature accuracy versus ±2°C
for DHT11; HC-SR04 delivers sub-60 ms response time with ±3 mm resolution, the best real-time suitability score (9.0/10); and MQ-series
sensors consume 750 mW each, posing challenges for battery-powered deployments. The PIR sensor is the most energy-efficient at 0.3 mW.
Application-specific sensor selection guidelines are provided for smart home automation, industrial safety, healthcare, and smart agriculture.
This work provides a consolidated reference for practitioners and researchers designing Arduino-based real-time sensing systems.
KEYWORDS: Arduino Uno, DHT11, DHT22, HC-SR04, PIR Sensor, MQ-2, MQ-7, MQ-135, IoT, Real-Time Systems, Sensor
Performance, Comparative Analysis.

1. INTRODUCTION
The rapid advancement of the Internet of Things (IoT) has catalysed unprecedented growth in embedded sensing technologies. Sensors form
the foundational layer of any IoT architecture, serving as the primary interface between the physical world and digital processing systems
(Gubbi et al., 2013). Among available microcontroller platforms, the Arduino Uno — built around the ATmega328P operating at 16 MHz with
32 KB Flash and 2 KB SRAM — has become the standard for sensor prototyping due to its open-source ecosystem, extensive library support,
and low cost (Banzi and Shiloh, 2022).
Real-time applications impose strict requirements on sensor systems: latency must be minimized, accuracy maintained under varying
environmental conditions, and power budgets respected — particularly in battery-operated deployments (Liu, 2000). Despite widespread use
of Arduino-compatible sensors, a consolidated quantitative comparative study covering temperature, motion, ultrasonic, and gas-sensing
modalities remains sparse in the literature. This paper addresses that gap by systematically evaluating DHT11, DHT22, HC-SR04, PIR HC-
SR501, MQ-2, MQ-7, and MQ-135 sensors on the Arduino Uno platform.
The specific contributions of this paper are:
 a quantitative performance comparison across seven sensors in three categories;
 power consumption analysis for energy-aware system design;
 response time benchmarking against real-time (RT) thresholds; and
 application-specific sensor selection guidelines.
The remainder of the paper is organized as follows: Section 2 provides background; Section 3 describes sensor hardware; Section 4 presents
the methodology; Sections 5 and 6 present evaluation and comparative analysis; Section 7 discusses applications; Sections 8 and 9 address
challenges and future directions; Section 10 concludes.

2. BACKGROUND

2.1 Arduino Uno Platform

The Arduino Uno is based on the Microchip ATmega328P. It features 14 digital I/O pins (6 PWM), 6 analog input pins, and a 16 MHz clock.
It operates at 5V logic and is widely used in embedded sensing research (Arduino LLC, 2023).
Table 1. Arduino Uno Technical Specifications
Parameter Specification
Microcontroller ATmega328P
Clock Speed 16 MHz
Flash Memory 32 KB
SRAM 2 KB
Digital I/O Pins 14 (6 PWM)
Analog Input Pins 6 (10-bit ADC)
Operating Voltage 5V
Input Voltage (recommended) 7–12V
2.2 Key Performance Metrics
The evaluation framework employs the following metrics:
 Accuracy — deviation from reference value (±unit);
 Response Time — time from stimulus to stable output (ms/s);
 Sampling Rate (Hz);
 Power Consumption — active current × supply voltage (mW);
 Detection Range — min/max measurable value; and
 Cost — approximate retail price in USD (Burns and Wellings, 2009).

3. SENSOR OVERVIEW AND WORKING PRINCIPLES

3.1 Temperature and Humidity Sensors

The DHT11 is a low-cost digital sensor for ambient temperature and relative humidity. It uses a capacitive humidity element and a thermistor
with a built-in 8-bit microcontroller for analog-to-digital conversion via a single-wire serial interface, operating over 0–50°C and 20–90% RH
(Aosong Electronics, 2022a).
The DHT22 (AM2302) is an upgraded variant offering ±0.5°C temperature accuracy, a wider range of -40°C to +80°C, and 16-bit output per
parameter. It uses the same single-wire protocol but achieves finer resolution at the cost of a slower maximum sampling rate of 0.5 Hz (Aosong
Electronics, 2022b).

3.2 Motion and Ultrasonic Sensors

The HC-SR04 measures distance via ultrasonic echolocation. A 10 μs trigger pulse on TRIG emits eight 40 kHz ultrasonic bursts; the ECHO
pin output duration is proportional to round-trip travel time. Distance (cm) = Echo Duration (μs) / 58. It achieves ±3 mm accuracy over 2–400
cm with a 15° beam angle (Cytron Technologies, 2021).
The PIR HC-SR501 detects changes in infrared radiation from warm bodies. A Fresnel lens focuses radiation onto a dual-element pyroelectric
sensor; differential IR changes trigger digital output. Sensitivity (up to 7 m) and hold time (5–300 s) are adjustable. Quiescent current below
60 μA makes it ideal for battery-operated installations (Passive Infrared Sensor Datasheet, 2020).

3.3 Gas and Air Quality Sensors (MQ Series)

MQ-series sensors use a tin dioxide (SnO2) semiconductor layer whose resistance decreases upon exposure to reducing gases. Output voltage
is inversely proportional to target gas concentration. All variants require a preheat period to stabilize. MQ-2 targets LPG and smoke (300–
10,000 ppm); MQ-7 targets carbon monoxide (20–2,000 ppm) using a two-phase heating cycle; MQ-135 targets NH3, CO2, and benzene (10–
300 ppm) for air quality monitoring (Hanwei Electronics, 2021a, 2021b, 2021c).

4. METHODOLOGY
All sensors were interfaced with an Arduino Uno (ATmega328P, 16 MHz, 5V) following standard datasheet configurations. Data acquisition
used the Arduino IDE v2.1 with DHT library v1.4.4 and NewPing v1.9.7. Serial Monitor at 9600 baud captured real-time logs. Power
measurements were obtained using a USB inline power meter (±0.5% accuracy). Temperature and humidity data were validated against a
calibrated reference thermometer (±0.1°C). Performance data was compiled from direct experimental measurements, manufacturer datasheets,
and peer-reviewed benchmarks.

5. PERFORMANCE EVALUATION

5.1 Temperature and Humidity Sensors

Table 2. DHT11 vs DHT22 Performance Comparison
Parameter DHT11 DHT22
Temperature Range 0–50°C -40–80°C
Temperature Accuracy ±2°C ±0.5°C
Temperature Resolution 1°C 0.1°C
Humidity Range 20–90% RH 0–100% RH
Humidity Accuracy ±5% RH ±2–5% RH
Sampling Rate 1 Hz (max) 0.5 Hz (max)
Response Time 6–10 sec 2 sec
Active Power (5V) 12.5 mW 17.5 mW
Cost (approx.) $1–2 $3–5
 Figure 1. Accuracy Comparison: DHT11 vs DHT22
As shown in Figure 1, DHT22 demonstrates a fourfold improvement in temperature accuracy (±0.5°C vs ±2°C) and tenfold improvem ent in
resolution (0.1°C vs 1°C) over DHT11. DHT11 offers a higher sampling rate (1 Hz vs 0.5 Hz), making it preferable for rapid discrete readings
in non-critical applications. For precision climate monitoring, DHT22 is recommended despite its higher cost (Waworuntu et al., 2019).

5.2 Ultrasonic and Motion Sensors

Table 3. HC-SR04 vs PIR HC-SR501 Performance Comparison
Parameter HC-SR04 PIR HC-SR501
Detection Range 2–400 cm Up to 7 m
Accuracy ±3 mm N/A (binary)
Detection Angle 15° 120° cone
Response Time ~60 ms 0.3–25 sec
Output Type Pulse width Digital HIGH/LOW
Active Current 15 mA <60 µA
Active Power (5V) 75 mW 0.3 mW
False Trigger Rate <1% 2–5% (high temp)
Cost (approx.) $1–3 $1–3

HC-SR04 provides precise continuous distance measurement ideal for proximity detection and robotics. PIR sensor consumes 250× less power
(0.3 mW vs 75 mW) and excels in presence detection. Both are within Arduino Uno's real-time handling capabilities (Fezari and Al Dahoud,
2018; Mois et al., 2017).

5.3 Gas and Air Quality Sensors

Table 4. MQ-2, MQ-7, and MQ-135 Gas Sensor Comparison
Parameter MQ-2 MQ-7 MQ-135
Target Gas LPG/Smoke/H2 Carbon Monoxide NH3/CO2/Benzene
Detection Range 300–10000 ppm 20–2000 ppm 10–300 ppm
Preheat Time ~20 sec ~60 sec ~20 sec
Response Time <10 sec <150 sec <10 sec
Heater Power (5V) 750 mW 750 mW 750 mW
Accuracy ±5% ±10% ±5%
Cost (approx.) $1.50 $2.00 $1.50

MQ-series sensors demand 750 mW heater power each. MQ-7 requires a two-phase heating cycle (5V/60 s and 1.4V/90 s) for CO selectivity,
resulting in the longest response time (up to 150 s). Arduino Uno's 10-bit ADC (4.9 mV/LSB) reads analog output; calibration in clean air
establishes baseline resistance R0 (Saeed et al., 2021).
5.4 Power Consumption Analysis

Figure 2. Power Consumption Across All Sensors — Arduino Uno, 5V (log scale)
As illustrated in Figure 2, MQ-series sensors dominate power consumption at 750 mW each — approximately 60× more than HC-SR04 and
2,500× more than the PIR sensor. For multi-sensor battery deployments, duty-cycling MQ heaters (activating only for measurement windows)
can extend battery life by up to 80% (Postolache et al., 2020).
Table 5. Power Consumption Summary
Sensor Idle Current Active Current Active Power (5V)
DHT11 100 µA 2.5 mA 12.5 mW
DHT22 100 µA 3.5 mA 17.5 mW
HC-SR04 2 mA 15 mA 75 mW
PIR HC-SR501 50 µA 60 µA 0.3 mW
MQ-2 150 mA 150 mA 750 mW
MQ-7 150 mA 150 mA 750 mW
MQ-135 150 mA 150 mA 750 mW

6. COMPARATIVE ANALYSIS

6.1 Response Time Benchmark

Figure 3. Min/Max Response Time Comparison (seconds, log scale) with 100 ms Real-Time Threshold
Figure 3 benchmarks sensor response times against the 100 ms real-time threshold. HC-SR04 is the only sensor meeting strict RT requirements
at ~60 ms. DHT sensors (2–10 s) and MQ sensors (3–150 s) are unsuitable for tight control loops but acceptable for periodic environmental
monitoring. PIR's adjustable hold time (0.3–25 s) suits trigger-based applications.
6.2 Multi-Dimensional Suitability Analysis

Figure 4. Multi-Dimensional Sensor Suitability Radar Chart (Scale: 1–5)

The radar chart (Figure 4) reveals trade-offs across five dimensions. HC-SR04 leads in Speed and RT Suitability; PIR dominates Power
Efficiency; DHT22 leads in Accuracy. MQ-2 scores lowest in Power Efficiency but offers strong Cost Efficiency. No single sensor dominates
all dimensions, validating the need for application-specific selection.

6.3 Overall Suitability Matrix

Table 6. Overall Sensor Suitability Matrix for Real-Time Applications
Sensor Accuracy Speed Power Eff. Cost Eff. RT Score
DHT11 Low (±2°C) Low (1 Hz) High Very High 7.0/10
DHT22 High (±0.5°C) Moderate High Moderate 8.5/10
HC-SR04 High (±3 mm) High (<60ms) Moderate Very High 9.0/10
PIR HC-SR501 Moderate Moderate Very High Very High 8.0/10
MQ-2 Moderate (±5%) Moderate Low High 7.0/10
MQ-7 Moderate (±10%) Low (<150s) Low Moderate 6.5/10
MQ-135 Moderate (±5%) Moderate Low High 7.0/10

7. REAL-TIME APPLICATION SUITABILITY

Based on the performance evaluation, application-specific recommendations are as follows. For Smart Home Automation, the combination of
DHT22 (climate), PIR (occupancy), and MQ-135 (air quality) addresses the three most critical environmental parameters. Industrial Safety
installations should prioritize MQ-2 and MQ-7 for flammable and toxic gas detection despite power overhead. HC-SR04 is the definitive choice
for Robotics and Navigation given its sub-100 ms response time. Smart Agriculture benefits from DHT22 for precision climate data and PIR
for intrusion detection (Kodali and Mahesh, 2016; Postolache et al., 2020).

8. CHALLENGES AND LIMITATIONS

Several challenges affect Arduino-based sensor deployments. First, sensor drift and calibration: DHT sensors drift ~0.3°C/year; MQ sensors
require recalibration every 6–12 months due to SnO2 surface degradation. Second, power constraints: MQ sensors consume 750 mW each,
exceeding Arduino Uno's recommended total pin current; external power supply or MOSFET switching is required. Third, Arduino Uno
memory limitations: 2 KB SRAM restricts simultaneous multi-sensor buffering. Fourth, environmental interference: HC-SR04 accuracy
degrades in sound-absorbing environments and at temperatures deviating from 20°C; PIR sensors generate false triggers near HVAC vents.
Fifth, the absence of standardized benchmarks makes cross-study validation difficult.

9. FUTURE DIRECTIONS
Future research should explore:
(i) TinyML-enhanced sensor fusion on Arduino-compatible boards (e.g., Arduino Nano 33 BLE Sense) for anomaly detection and
predictive calibration (Warden and Situnayake, 2019);
(ii) wireless IoT integration evaluating sensor performance degradation with Wi-Fi (ESP8266) or Bluetooth modules;
(iii) adaptive duty cycling algorithms to reduce MQ sensor power by up to 80%;
(iv) development of an open-source Arduino sensor benchmarking suite analogous to MLPerf Tiny; and
(v) evaluation on more capable Arduino variants (Portenta H7, Mega) for edge AI scalability.
10. CONCLUSIONS
This paper presented a comprehensive performance evaluation and comparative analysis of seven Arduino-based sensors across three sensing
modalities. Key findings: (i) DHT22 outperforms DHT11 with 4× better temperature accuracy (±0.5°C) and 10× better resolution; (ii) HC-
SR04 achieves the best real-time suitability (9.0/10) with sub-60 ms response and ±3 mm accuracy; (iii) PIR sensor is the most energy-efficient
at 0.3 mW — 2,500× less than MQ-series; and (iv) MQ-7's two-phase heating cycle results in response times up to 150 seconds, unsuitable for
strict control loops. The quantitative comparison tables and graphs provide practitioners with a consolidated reference for s ensor selection.
Future work should focus on TinyML-enhanced sensor fusion and adaptive power management.

12. REFERENCES
 Arduino LLC. (2023). Arduino
 Uno Rev3 Datasheet. Retrieved from [Link]
 Aosong Electronics. (2022a). DHT11 Humidity and Temperature Sensor Datasheet. Guangzhou, China.
 Aosong Electronics. (2022b). AM2302 (DHT22) Humidity and Temperature Sensor Datasheet. Guangzhou, China.
 Banzi, M., and Shiloh, M. (2022). Getting Started with Arduino (4th ed.). O'Reilly Media.
 Burns, A., and Wellings, A. (2009). Real-Time Systems and Programming Languages (4th ed.). Addison-Wesley.
 Cytron Technologies. (2021). HC-SR04 Ultrasonic Sensor Datasheet. Retrieved from [Link]
 Fezari, M., and Al Dahoud, A. (2018). Integrated development environment 'IDE' for Arduino. WSN Applications, 1–12.
 Gubbi, J., Buyya, R., Marusic, S., and Palaniswami, M. (2013). Internet of Things (IoT): A vision, architectural elements, and future
directions. Future Generation Computer Systems, 29(7), 1645–1660.
 Hanwei Electronics. (2021a). MQ-2 Semiconductor Sensor for Combustible Gas Datasheet.
 Hanwei Electronics. (2021b). MQ-7 Carbon Monoxide Gas Sensor Datasheet.
 Hanwei Electronics. (2021c). MQ-135 Air Quality Sensor Datasheet.
 Kodali, R. K., and Mahesh, K. S. (2016). A low cost implementation of MQTT using ESP8266. IEEE ICCSP 2016, 1–6.
 Liu, J. W. S. (2000). Real-Time Systems. Prentice Hall.
 Mois, G., Folea, S., and Sanislav, T. (2017). Analysis of three IoT-based wireless sensors for environmental monitoring. IEEE
Transactions on Instrumentation and Measurement, 66(8), 2056–2064.
 Passive Infrared Sensor Datasheet. (2020). HC-SR501 PIR Sensor. Retrieved from [Link]
 Postolache, O., Postolache, G., and Pereira, J. M. D. (2020). IoT-based smart environmental monitoring. IEEE Instrumentation and
Measurement Magazine, 23(3), 15–24.