Optimizing Residential Energy Usage with Smart Devices: A Case Study on Energy Efficiency and Environmental Sustainability

Abstract

In this study, we examined the impact of an intelligent system and air conditioning control on power consumption. The experiment was carried out during five distinct time periods: (1) background room usage, (2) smart system setup, (3) air conditioning control to maintain room temperature at no more than 27 °C, (4) air conditioning temperature control during working hours, and (5) air conditioning operated continuously to maintain the room temperature at 27 °C. For each time period, the daily power consumption was evaluated, and outliers were identified and eliminated using a threshold derived from the hourly average. The findings demonstrated that the smart system setup period and air conditioning control resulted in lower usage compared to continuously operated air conditioning with substantial spikes in demand. The impacts of the novel system and air conditioning control on energy consumption were revealed through statistical analysis, which included regression models and hypothesis tests. According to this study’s findings, it is essential to regulate spikes and guarantee proper operation to reduce the carbon footprint while maintaining a comfortable atmosphere. Notably, the integration of the smart system and optimized scheduling resulted in a substantial decrease in greenhouse gas emissions, with annual carbon emissions reduced by up to 65% compared to continuously operated air conditioning without smart control. Moreover, these systems can optimize energy use.

1. Introduction

The growing demand for energy-efficient systems is critical in reducing carbon footprints and achieving sustainable environmental goals. As global energy consumption rises, with buildings being significant energy consumers and contributing substantially to climate change [1], particularly in residential and commercial buildings, it is essential to explore technologies that can optimize electricity usage while maintaining comfort levels [2]. Intelligent systems and air conditioning (A/C) control technologies are emerging solutions to improve energy efficiency by automatically regulating temperature, humidity, and overall energy consumption. Recently, many researchers have investigated the integration of smart meters into energy management systems in standard rooms to increase energy efficiency and user convenience. Modifications of smart systems for various applications have been proposed in Thailand [3,4,5]. The advancement of smart systems has revolutionized multiple sectors by integrating automation, artificial intelligence, and IoT (Internet of Things) technologies to enhance efficiency, convenience, and sustainability. Smart systems are designed to intelligently monitor, analyze, and control physical environments, enabling adaptive responses to dynamic conditions. One of the most significant applications of smart systems is in the domain of building management, where optimizing energy consumption and improving indoor comfort are key objectives. Air conditioning (A/C) systems, which are critical for maintaining indoor thermal comfort, are among the largest consumers of energy in residential, commercial, and industrial buildings, and their consumption tends to increase as the global temperature escalates and the frequency of heat waves increases [6].

Smart systems, specifically air conditioning (A/C) control technologies, are emerging as crucial solutions to improve energy efficiency by automatically regulating temperature, humidity, and overall energy consumption in buildings [7,8]. The implementation of such intelligent control systems promises significant benefits, including reduced operational costs, minimized power losses, improved power quality, and increased capacity to accommodate distributed and renewable energy sources, thereby contributing to a more sustainable and secure energy future. Traditional A/C control methods, often based on fixed schedules or manual operation, are generally inefficient and unable to respond effectively to changing environmental and occupancy conditions [9]. Consequently, this has led to substantial energy wastage, increased operational costs, and adverse environmental impacts, including higher greenhouse gas (GHG) emissions. To address these challenges, researchers and industry professionals have increasingly focused on developing intelligent A/C control technologies that leverage smart system capabilities. These smart A/C control systems utilize sophisticated sensors, machine learning algorithms, and network connectivity to enable real-time monitoring and adaptive control [10,11,12]. By analyzing variables such as indoor temperature, humidity, occupancy patterns, outdoor weather conditions, and user preferences, these systems can dynamically adjust the operation of air conditioning units to optimize energy use without compromising comfort [13,14].

In addition, the integration of IoT devices has facilitated remote control and data-driven decision-making through cloud computing platforms. Smart thermostats and building automation systems exemplify practical implementations that have demonstrated significant energy savings and improved user experience. Recent studies have demonstrated the effectiveness of advanced machine learning and optimization techniques—including deep learning, reinforcement learning, and hybrid models—in predicting energy demand, optimizing HVAC operations, and improving the overall efficiency of smart homes and buildings [14,15]. These approaches further highlight the growing potential of data-driven and adaptive control strategies to enhance the performance and sustainability of residential energy systems. These systems are designed to optimize energy consumption while ensuring user comfort and satisfaction. Some of them may include technologies such as machine learning [16] and genetic algorithms to predict or adjust to energy usage patterns more effectively [17], and others involve the implementation of intelligent energy optimization in smart homes, which utilizes smart systems for configuring energy consumption [18,19].

Furthermore, the emergence of renewable energy integration and smart grids presents new opportunities for smart A/C control systems to contribute to broader energy management strategies. By aligning cooling demand with renewable energy availability or grid requirements, these technologies support demand response programs and enhance the sustainability of building operations. In spite of the promising advancements, challenges remain in developing robust, scalable, and user-friendly smart A/C control systems. Issues such as system complexity, cost, interoperability, privacy concerns, and user acceptance need to be addressed to ensure widespread adoption [20].

Information for evaluating the effectiveness of an intelligent system-controlled environment has been collected through various studies [21,22]. On the other hand, a specific and significant challenge in designing energy-efficient systems, such as smart A/C controls, is maintaining comfortable environmental levels that align with established comfort standards. ASHRAE Standard 55 (2020) [23] defines a comfortable thermal range for indoor environments, typically between 20 °C and 27 °C, with relative humidity levels between 40% and 65%. These parameters are essential for ensuring people stay well without excessive energy use. Thus, achieving both optimal indoor thermal comfort and effective energy conservation presents a significant challenge, as these objectives often require careful trade-offs in building control strategies. While smart system design has seen numerous advancements over the past several years, the primary challenge remains in developing and implementing these systems to effectively comprehend and respond to user behaviors and preferences [24]. Without this crucial user-centric approach, smart systems may risk not achieving their full potential and sustainability. This study aims to investigate how smart system electricity control can optimize energy consumption and user experience in residential settings and evaluate the impact on GHG emission reduction, with a particular focus on different electricity consumption patterns under various conditions, and this study also aims to determine the principal concept of smart system electricity control. In addition, we carefully examine the current challenges, such as balancing comfort and energy efficiency, and highlight potential areas for future research and development, aiming to contribute to a more efficient, reliable, and sustainable global energy system.

2. Materials and Methods

2.1. The Smart System and Experimental Design

This experiment was conducted in a meeting room at U-Media (University Academic Service Center: UNISERV, Chiang Mai University, Thailand) during the summer season from 22 April to 13 May and 1 to 2 June 2025. Initially, the room was a standard meeting space with conventional lighting controlled by switches and was modified to include a smart control system to regulate electricity usage. Before implementing the smart system, a Wi-Fi access system was installed to enable remote control and data collection via the cloud, facilitating data analysis after the experiment. The electricity usage of the room was recorded during the background phase, where no equipment was installed, to serve as a baseline for comparison with subsequent data. The test period was divided into five stages as follows:

Period 1—Background Data (22 to 26 April 2025): Electricity usage data from the room were recorded before any system was installed.

Period 2—Smart System Installation (2 May to 4 May 2025): After the smart system was fully installed, electricity usage was recorded for three consecutive days.

Period 3—Temperature Control (6 to 8 May 2025): During this phase, the room’s temperature was controlled to remain below 27 °C with a relative humidity (RH) range of 40–60%, in compliance with ASHRAE Standard 55. The air conditioning system was operated continuously for 24 h a day for three days, ensuring the room remained comfortably cool.

Period 4—Working-Hour Temperature Control (9 to 11 May 2025): This period followed the same conditions as Period 3, but the air conditioning was limited to working hours (08:00–17:00).

Period 5—The air conditioning system was operated for 24 h without temperature control (1 to 2 June 2025) to compare electricity usage with Periods 3 and 4, respectively.

For both temperature control periods, the air conditioner’s fan was set to run continuously to maintain humidity below 65%, while the compressor was activated only when the room temperature exceeded 27 °C. The compressor followed a 7 min on, 3 min off cycle until the temperature dropped below 27 °C.

The devices used in this experiment are shown in

Table 1. Sensor properties.

Device	Model/Type	Key Specifications	Accuracy
Electricity Meter	Acrel ADL200	AC220V, 10(80)A, Class 1.0 accuracy, RS485 MODBUS-RTU, LCD, 90 × 36 × 65 mm, <10 VA (voltage), <4 VA (current)	kWh Accuracy: Class 1.0 (±1%)
Temperature and Humidity Sensor (in testing room)	Xiaomi LYWSD03MMC	Temp: 0–60 °C, Humidity: 0–99% RH, Bluetooth 4.2 BLE, CR2032 battery, 43 × 43 × 12.5 mm	Temp: ±0.1 °C; Humidity: ±1% RH
Temperature and Humidity Sensor (Outdoor measurement)	Industrial Grade RS485 Temperature, SHT31	Temp: −40–60 °C, Humidity: 0–100%	Temp: ±0.3 °C; Humidity: ±0.8% RH
Motion Sensor	Xiao-ESP32-C3 + IR/mmWave	ESP32-C3, WiFi/BLE, mmWave 24 GHz, Range: 0–5 m, Detection angle: 90°H/60°V, OTA support	Not specified
Air Controller	Grove 2-Channel SPDT Relay	2 SPDT, 5 V DC control, Max: 250 V AC/10 A, 110 V DC/10 A, Power: ~0.45 W, 40 × 40 × 19 mm	Contact Resistance: ≤100 mΩ
Light Controller	Grove 4-Channel SPDT Relay	4 SPDT, 5 V DC control, Max: 250 V AC/10 A, 110 V DC/10 A, STM32F030F4P6 controller, I2C/SWD, 30 ops/min	Contact Resistance: ≤100 mΩ
Wi-Fi Router Repeater	TP-LINK, model Archer C54	AC1200 Dual Band Wi-Fi Router speeds up to 1200 Mbps, Multi-Mode: Router, Access Point, Range Extender	Not specified
Control and Monitoring Software	Home Assistant	Open-source software for home automation; in this work, ESPHome was used for system integration	Not specified

.

All sensors were calibrated, and communication pathways were verified. Following calibration, predefined control rules were established, particularly triggering the A/C unit’s activation when temperatures exceeded 27 °C or occupancy was detected.

The system workflow for smart control and monitoring is shown in Figure 1. An electricity meter (Acrel ADL200, Jiansu Acrel Electrical Manufacturing Co., Ltd., Jiangyin, China) was installed to provide the real-time monitoring of the room’s total electricity consumption, with data transmitted to a central processing unit for analysis. For environmental monitoring, a Xiaomi LYWSD03MMC sensor, manufactured by Xiaomi in Beijing, China, measured indoor temperature and humidity, while another temperature and humidity sensor (Sensirion model SHT31, Sensirion AG, Stäfa, Switzerland) recorded outdoor conditions. These sensors provided continuous feedback to the control system, ensuring that the indoor environment remained within the ASHRAE comfort range of 20–27 °C and 40–65% relative humidity. Occupancy detection was managed by a Xiao-ESP32-C3 microcontroller (manufactured by Seeed Studio in Shenzhen, China) equipped with an infrared motion sensor positioned to detect movement within the room. This allowed the system to dynamically adjust energy consumption based on real-time occupancy status, such as activating the air conditioning system only when motion was detected (e.g., activating cooling systems when motion was detected). The air conditioning system was controlled via a Grove 2-Channel SPDT Relay (manufactured by Seeed Studio in Shenzhen, China), enabling automated switching based on sensor input. The relay received commands from the central control logic to maintain the desired temperature and humidity. Lighting was similarly managed through a Grove 4-Channel SPDT Relay (manufactured by Seeed Studio in Shenzhen, China), allowing for the remote and automated control of the room’s lights. All devices communicated through a network access point (Wi-Fi router repeater, TP-LINK model Archer C54, manufactured by TP-Link Technologies Co., Ltd. In Shenzhen, Guangdong, China), ensuring reliable data transfer between sensors, controllers, and the cloud system. Home Assistant software (version 2025.4.1, an open-source sponsored by Nabu Casa in Irvine, CA, USA), which is integrated with the system, is responsible for monitoring, controlling, and automating the devices. It allows users to set rules, download data, and remotely monitor the system’s performance via an IoT platform. The IoT cloud was integrated to enable remote access to the system for monitoring and control. Users can use the cloud to adjust settings such as temperature thresholds, motion detection rules, and electricity usage limits. All devices continuously send real-time data (e.g., electricity usage, temperature, humidity, and motion detection) to the Home Assistant via the access point.

2.2. Air Conditioning Control Cycle Design

In this study, we developed a smart air conditioning (A/C) control system to optimize both energy efficiency and occupant comfort in accordance with the guidelines of ASHRAE Standard 55. The control cycle was designed so that both the fan and compressor operate together for 7 min, after which the compressor turns off for 3 min while the fan continues to run. This 7 min on, 3 min off pattern was established based on preliminary experiments assessing temperature stability and energy consumption in the test room. This approach ensures that the indoor temperature remains within the ASHRAE comfort range (20–27 °C) and that room ventilation is maintained. The choice of a 3 min compressor off period is critical: if the compressor is off for more than 3 min, the room temperature rises too quickly, failing to meet the ASHRAE standard.

On the other hand, turning the compressor back on sooner than 3 min increases its workload, leading to higher electricity consumption and unnecessary mechanical stress on the system. This control strategy was validated in real time by monitoring temperature and humidity. The results showed that a 3 min compressor off period is optimal for preventing the temperature from exceeding 27 °C while also reducing excessive compressor cycling and overall energy use. The continuous operation of the fan during the compressor’s rest period helps maintain air circulation and humidity control, supporting comfort requirements. The smart control system automates this cycle using real-time data from temperature and humidity sensors. The compressor is activated when the room temperature exceeds 27 °C and operates according to a 7/3 min schedule. The fan remains on at all times to ensure even temperature distribution and to keep humidity below 60%, as recommended by ASHRAE. Users can remotely monitor and adjust system settings via the Home Assistant interface. All operational data (temperature, humidity, electricity use, and motion detection) is stored in the cloud for further analysis and optimization. To minimize interference from external sources, the tests were conducted when the room was empty. Utilizing the information gathered throughout these stages, an evaluation was conducted to determine whether it would be possible to successfully adopt a novel system that would reduce power consumption while maintaining optimal environmental conditions in the room. This experiment aimed to evaluate the possibility of intelligent systems optimizing energy use without requiring any manual interaction, thereby minimizing the time consumers have to spend adjusting temperature settings manually. We also investigated the problems and obstacles associated with implementing such a system in a conventional room. Additionally, this study provides a platform for future tests to be conducted in rooms frequently used by occupants.

3. Results

3.1. The Electricity Consumption

The experiment’s findings are presented below, including information on the amount of power used, temperature, and humidity levels observed during the testing periods. During these phases, an evaluation of the performance of a smart system and air conditioning (A/C) control was conducted under various settings. The primary purpose of this study was to determine whether or not these systems could lower the amount of power consumed while also maintaining a pleasant environment per the requirements set out by ASHRAE [25,26] for temperature and humidity.

From Figure 2, overall electricity consumption during the background period (a) was low, with occasional early-morning peaks (~06:00), likely due to brief activity such as lighting or equipment startup. Consumption remained below 0.017 kWh, consistent with the absence of active systems. The temperature remained steady at around 28–30 °C, while humidity fluctuated between 49% and 70%, within or slightly above the ASHRAE-recommended levels. After the smart system was installed (b), electricity usage increased somewhat and exhibited a more structured, cyclical pattern, reflecting system modulation based on environmental needs. Peaks stayed below 0.018 kWh, suggesting that energy was only used when necessary. The temperature was tightly regulated at around 28–29 °C, which was more stable than in the background phase. Humidity rose slightly (average ~70%). When 24 h temperature control (c) was implemented, a notable increase was observed. Electricity consumption peaked at approximately 0.4 kWh during daytime hours, and the continuous usage throughout the 24 h cycle reflects the constant operation of the air conditioning system. The temperature was successfully maintained between 25 and 26 °C, meeting ASHRAE standards. Humidity dropped slightly (~63%), showing improved environmental control. As the A/C was operated only during working hours (d), it was found that electricity usage spiked during working hours (08:00–17:00) and dropped to near zero at other times, indicating the successful implementation of a time-based control strategy. The temperature remained stable, maintained at just below 27 °C during working hours and slightly higher afterward, remaining within comfortable limits. Humidity fell within the 50–75% range and remained within acceptable bounds. During the A/C 24 h period without smart temperature control (e), as shown in the figure, electricity usage remained consistently high throughout the day and night, averaging approximately 0.29 kWh per hour, indicating the continuous operation of the air conditioning system regardless of demand or occupancy. The indoor temperature was maintained at a stable level of approximately 25 °C, and the relative humidity remained steady at around 60–65%, indicating effective but energy-intensive cooling and dehumidification. This operational mode, which serves as the baseline in this study, demonstrates the high energy cost of running the A/C continuously without any adaptive control, resulting in the maximum electricity consumption and stable comfort conditions. However, it also highlights significant potential for energy and carbon savings if smart or scheduled control strategies are applied.

Table 2. Hourly averaged electricity consumption *, temperature, and humidity across testing periods.

Period	Indoor Measurement (n ≥ 46)			Outdoor Measurement (n ≥ 46)
	Electricity (Kwh)	Temperature (°C)	Humidity (%)	Temperature (°C)	Humidity (%)
Background (22–26 April)	0.017 ± 0.020	29.78 ± 0.63	64.96 ± 3.83	38.48 ± 0.95	67.94 ± 8.91
Smart System (2–4 May)	0.018 ± 0.015	28.76 ± 0.43	71.39 ± 4.74	34.60 ± 1.05	81.06 ± 5.99
A/C 24 h (6–8 May)	0.168 ± 0.094	27.31 ± 0.73	64.26 ± 5.27	37.03 ± 1.02	85.03 ± 7.69
A/C Working Hours Only (9–11 May)	0.099 ± 0.102	28.08 ± 1.07	65.96 ± 7.22	37.65 ± 1.36	79.36 ± 7.95
A/C Working Without Temperature Control (1–2 June)	0.292 ± 0.045	25.42 ± 0.27	62.82 ± 1.18	34.65 ± 0.67	78.83 ± 5.29

* Considering every hour of the tested period regardless of whether the smart system is working or not.

presents a report that provides an overview of the overall results.

The background period, where the room operated under normal conditions without any smart system or air conditioning control, provided a baseline for comparison. The average hourly electricity usage during this period was 0.017 kWh per hour. The average temperature was 29.78 °C, and the average humidity was 64.96%. During the smart system setup phase, the average hourly electricity usage increased slightly to 0.018 kWh per hour. The temperature decreased marginally to 28.76 °C, and humidity increased to 71.39%. This period indicates that while the smart system was operational, it did not significantly reduce electricity consumption compared to the background period. The increase in humidity suggests that the smart system may not have effectively controlled moisture levels during this phase. It is noted that the baseline represents no activity in the room, to prove that after installing smart system devices, the electricity consumption does not increase. To estimate energy optimization, consider Period 5 as the baseline instead. In the A/C control phase, where the room’s temperature was kept below 27 °C at all times, electricity consumption increased substantially to an average of 0.168 kWh per hour. The temperature was maintained at an average of 27.31 °C, and humidity was 64.26%, indicating that the cooling system effectively maintained comfort. However, the high electricity usage reveals the energy-intensive nature of continuous air conditioning [25]. During the working hour-only A/C control period, where air conditioning was limited to 08:00–17:00, electricity usage was lowered, averaging 0.099 kWh per hour. Thus, limiting A/C operation to working hours, this period exhibited lower energy consumption than the 24 h control phase. The temperature was maintained at an average of 28.08 °C, and humidity was 65.96%, which is still within acceptable comfort levels. However, the inefficiency of the cooling system resulted in excessive energy use during working hours.

The experimental data revealed distinct differences in daily electricity consumption across different periods, as shown in Figure 3. The trend in daily electricity usage aligned with the hourly consumption patterns. Specifically, the baseline energy consumption, used as a reference, was measured at 0.424 kWh per day. During the operation of the smart control system, the energy consumption increased slightly to 0.511 kWh/day. In contrast, the continuous 24 h operation of the air conditioning system resulted in significantly higher consumption, at 3.972 kWh/day, while scheduled A/C operation during working hours consumed 2.367 kWh/day. Despite a small increase over the background period, the smart system achieves approximately 15% energy savings compared to continuous A/C operation without temperature control and around 49% savings compared to scheduled working-hour operation. This demonstrates that adaptive, occupancy-based smart control systems can outperform traditional methods for energy efficiency and operational flexibility.

3.2. Statistical Analysis

Statistical tests were performed to assess the significance of the observed differences in electricity usage, temperature, and humidity.

• Two-Sample t-Test for Comparing Background vs. Smart System

We performed a t-test to determine whether the difference in electricity usage between the background and smart system periods was statistically significant.

Null hypothesis ($H_0$): There is no significant difference in electricity usage between the background and smart system periods.

$$H_0 :{\mu }_{background }={\mu }_{smart system}$$

Alternative hypothesis ($H_A$): There is a significant difference in electricity usage between the background and smart system periods.

$$H_A :{\mu }_{background }\ne {\mu }_{smart system}$$

The t-statistic for the two-sample t-test is calculated as follows:

$$t={\displaystyle \frac{\overline{x}1-\overline{x}2}{\sqrt{\frac{s_1^2}{n_1}}+\frac{s_2^2}{n_2}}}$$

where

$\overline{x}1$ and $\overline{x}2$ are the sample means for the two periods (background and smart system).

$s_1^2$ and $s_2^2$ are the sample variances.

$n_1$ and $n_2$ are the sample sizes.

The t-statistic helps determine whether the difference between the two periods is statistically significant by comparing it to a critical value from the t-distribution or using a p-value.

2.

One-way ANOVA for Comparing All Five Phases

To compare the mean electricity usage across all five periods (background, smart system, A/C control 24 h, A/C control working hours, and A/C 24 h without temperature control), we performed a one-way ANOVA.

Null hypothesis ($H_0$): The means of electricity usage across all five periods are equal.

$$H_0 :{\mu }_{background }={\mu }_{smart system}={\mu }_{A/C24h}={\mu }_{A/Cworking hours}$$

Alternative hypothesis ($H_A$): At least one of the periods has a different mean electricity usage.

$H_A$: At least one mean is different.

The F-statistic for the ANOVA is calculated as follows:

$$F={\displaystyle \frac{Between-group variance}{Within-group variance}}$$

where

Between-group variance measures the extent to which the means of the different groups differ from the overall mean.

Within-group variance measures the extent to which data points within each group differ from their respective group means.

Suppose that the p-value associated with the F-statistic is less than 0.05. In that case, we reject the null hypothesis and conclude that there are significant differences in electricity usage among the different periods.

• Two-Sample t-Test (Background vs. Smart System)

A t-test was performed to compare electricity usage between the background and smart system installation periods. The results are as follows:

t-statistic: −2.234.

p-value (two-tail): 0.0355 < 0.001.

This indicates that the difference in electricity usage between the background and smart system periods is statistically significant. Therefore, the smart system resulted in a significant increase in electricity consumption compared to the baseline.

2.

One-Way ANOVA (All Periods)

An ANOVA test was conducted to compare electricity usage across all five periods (background, smart system, A/C 24 h (without temperature control), A/C 24 h (temperature control), and A/C working hours (temperature control)). The results are as follows:

F-statistic: 17.26.

p-value: <0.001.

The ANOVA test confirms significant differences in electricity usage across the five testing periods. Specifically, the A/C control periods (two conditions for A/C 24 h and one condition for working hours) resulted in substantially higher energy consumption than that in the background and smart system periods, as there was no activity during the experiment launch.

3.3. Estimation of GHG Emission Reduction

To evaluate the environmental impact of the smart system and air conditioning (A/C) control interventions, we estimated the reduction in greenhouse gas (GHG) emissions resulting from changes in electricity consumption across the various test periods. The standard formula for calculating annual greenhouse gas (GHG) emission reductions from energy efficiency projects, particularly in the context of the Clean Development Mechanism (CDM) and related international standards for carbon accounting, is shown below.

$$ERy=({EC}_{i,BL,y}\times {10}^{-3}\times {EF}_{EC,PJ,y})-({EC}_{PJ,y}\times {10}^{-3}\times {EF}_{EC,PJ,y})$$

where

${ER}_y$: Emission reductions in year y.

(typically expressed in tonnes CO₂ equivalent per year).

${EC}_{i,BL,y}$: Baseline energy consumption of energy type I in year y.

${EC}_{PJ,y}$: Project energy consumption in year y.

${ER}_{EC,PJ,y}$: Emission factor for the energy consumption of the project in year y.

10⁻³: Conversion factor from units (often kWh or similar) to tons.

Table 3. Comparison of carbon emission released with different air control systems.

Scenario No.	Air control systems			Reduce carbon emission (ton/year)
	Turn on air conditioning for 24 h without smart system (ton/year)	Turn on air conditioning for 24 h combined with smart system (ton/year)	Turn on air conditioning during working hours only and combined with smart system (ton/year)
1	1.277234	-	0.448419	0.828814
2	1.277234	0.704152	-	0.573082
3	-	0.704152	0.448419	0.255732

presents a comparative analysis of annual carbon emissions under different air conditioning (A/C) operational scenarios, both with and without the smart system. The results demonstrate the significant impact that smart control and optimized scheduling can have on reducing greenhouse gas (GHG) emissions. The highest carbon emissions are observed in this scenario, with 1.277234 tons of CO₂ per year. This reflects the energy-intensive nature of running the A/C continuously. Implementing the smart system reduces annual emissions to 0.704152 tons of CO₂, a reduction of approximately 0.573082 tons (about 45%) compared to continuous operation without smart controls. This improvement is attributed to the smart system’s ability to optimize runtime and environmental set points, even during continuous operation. Limiting A/C use to working hours and combining it with smart control achieves the lowest emissions, at 0.448419 tons of CO₂ per year. This represents a reduction of 0.828814 tons (about 65%) compared to the baseline of 24 h operation without smart control. The greatest GHG emission reduction is realized when both operational scheduling (working hours only) and smart system optimization are applied together. The data highlight that the strategic use of smart controls, especially when paired with occupancy-based or time-based scheduling, can significantly reduce the carbon footprint of building cooling operations.

4. Discussion

The electricity usage data reveals significant insights into the impact of smart systems [27,28] and air conditioning control on energy consumption [27]. The period during which the smart system was being set up (2 to 4 May 2025) saw a modest increase in power consumption compared to the background period (22 to 26 April 2025). The functioning of the smart system or additional devices added as part of the configuration may also be responsible for this rise. Compared to the background period, the temperature and humidity were more effectively managed, which suggests that the system had a role in regulating environmental parameters. As a result of the air conditioning being adjusted to maintain a temperature below 27 degrees Celsius continuously for twenty-four hours (from 6 to 8 May 2025), the amount of power consumed increased to an average of 0.168 kWh per hour. Due to this significant rise, the energy-intensive nature of continuous air conditioning is highlighted, resulting in a substantially larger power demand [28,29] compared to non-smart system control times for A/C. The system efficiently maintained a pleasant temperature and humidity level within acceptable ranges, despite consuming a significant amount of energy. The power usage was considerably lower (0.099 kWh) during the working hour-only A/C control period (9–11 May 2025).

Under real-world conditions in tropical climate characterized by elevated temperature and high humidity, air conditioning systems exhibit increased energy consumption to achieve and maintain thermal comfort within the preferred indoor environment. Based on the results, it appears that the air conditioning system during the working-hour phase may have been inefficient, meaning that it used more energy than was required to maintain the desired temperature and humidity levels. This may be connected to the ventilation system that is located in the testing room [30,31]. The temperature was maintained at around 28.25 degrees Celsius during working hours, and this was the only change made. Nevertheless, in comparison to continuous monitoring, it was not administered with the same degree of efficiency. A possible explanation for this is that the space was not well-regulated in terms of temperature, resulting in higher temperature (see Table 1) swings from the outside [6,32,33]. As a result, the temperature within the room would initially be higher at the beginning of working hours, in comparison to the typical scenario of continuous air conditioning management. The ineffectiveness of this technique is demonstrated by the fact that the temperature was maintained within a comfortable range despite requiring a significant amount of energy to achieve the desired temperature, as mentioned in earlier research [34]. However, further studies should include multi-room demonstration and long-term assessment to capture seasonal or occupancy variations.

The findings of the experiments also indicated that the smart control system developed is capable of efficiently reducing the amount of energy consumed while simultaneously maintaining the level of comfort inside the indoor environment in accordance with ASHRAE standards. The dependability of the sensor feedback loop and control algorithms is demonstrated by the fact that the system maintains the temperature and humidity within the required comfort range. Additionally, connectivity with Home Assistant and cloud-based platforms enables seamless remote monitoring and control, significantly enhancing the user experience and increasing the system’s versatility [35], which shows the potential of the system for widespread implementation and further development at a relatively low cost. However, the researchers were also confronted with several obstacles, including the difficulty of initial setup and calibration, concerns over confidentiality related to stable network connectivity, and the occasional occurrence of false triggers from motion sensors, as documented in earlier research [36,37,38,39]. Addressing these concerns in greater depth in further work is necessary to enhance the dependability of the system and encourage wider adoption.

5. Conclusions

According to the t-test, which had a p-value of less than 0.001, the smart system configuration resulted in a marginal increase in power consumption compared to the baseline. The fact that the 24 h A/C control period resulted in significant electricity usage (0.074 kWh per hour) exemplifies the energy-intensive nature of continuous air cooling. During the working hour-only air conditioning control period, the amount of power used was much greater (0.098 kWh per hour), indicating inefficiencies in the system’s cycling and cooling operation. The statistical tests (t-test and ANOVA) reveal substantial differences in the amount of electricity used during each of the periods, which lends credence to the conclusion that air conditioning control had a significant influence on the amount of energy consumed.

In addition, this work found the environmental value of integrating smart energy management systems in building operations. While continuous A/C use leads to the highest emissions, both smart control and schedule optimization independently and synergistically reduce energy consumption and associated GHG emissions. The results also suggest that even in scenarios where comfort must be maintained, substantial emission reductions are achievable without sacrificing indoor environmental quality. This evidence supports policy and practical recommendations to adopt smart control technologies and operational strategies as effective measures for climate change mitigation in the building sector. Future work could further enhance these benefits by incorporating real occupancy data, predictive algorithms, and integration with renewable energy sources.

While the proposed system demonstrates strong real-world performance and practical advantages, future work will consider the integration of advanced optimization algorithms, such as those highlighted in recent studies [40,41,42,43], to further enhance energy efficiency and flexibility in residential applications. Future recommendations might include incorporating real-time adaptive controls or occupancy-based cooling into systems in order to make them more energy-efficient. This would ensure that comfort levels are maintained while reducing energy usage. Additionally, enhancing the logic of smart systems and cooling cycles can contribute to achieving sustainability objectives by reducing the environmental footprint.