Samarth Rural Educational Institute

SAMARTH COLLEGE OF ENGINEERING,

BELHE

A Project Review On

A
SYNOPSIS REPORT ON
The framework for IoT based appliances
recognition in smart homes
Under Guidance Of
Prof. Borhade R. B.
By
Lamkhade Sanika Machhindra

Department of Computer Engineering

 Academic Year: -2024-25
Paper Title:-
The framework for IoT based appliances recognition in smart homes

Abstract :-
Internet of Things (IoT) technologies will play an important role in enabling the smart Grid
achieving its goals in monitoring, protecting, and controlling by incorporating sensors,
actuators, and Metering devices while supporting various network functions and system
automation. In this regard, home Energy management systems (HEMS) enable customers
efficiently use energy by managing their consumption, providing feedback information and
improving control of major appliances. This work proposes a Novel framework for IoT based
appliance recognition in smart homes. It consists of two parts: training Framework and
inference framework. The proposed framework allows incorporating different loads in the
Monitoring system and enables selecting and testing specific parameters related to dataset
configuration Feature extraction, and classifier model setting. The work contributes by
developing an easy-to-use tool That allows customization of the training/prediction
parameters according to the user criterion. Once the data and all its parameters are loaded, a
novel feature extraction algorithm is used to obtain a total of ten Statistical features. For the
classification task, three machine learning models are included: a feed-forward Neural
network, a long short-term memory and a support vector machine. In addition, the user can
apply a Set of techniques to handle imbalanced classes, and also measure the influence of the
selected features in the Classifiers’ prediction by performing a feature importance analysis.

Introduction :-
The advent of smart home technology has revolutionized the way energy is consumed and
managed in residential settings. With the proliferation of IoT devices, homeowners can now
monitor their energy usage in real-time, leading to more informed decisions regarding
consumption patterns. However, accurately identifying and monitoring individual appliances’
energy consumption remains a significant challenge. This research proposes a comprehensive
framework for IoT-based appliance recognition that leverages machine learning techniques to
enhance home energy management systems (HEMS). By accurately recognizing appliance
usage, the framework aims to provide users with actionable insights into their energy
consumption, ultimately promoting energy efficiency and sustainability.
Smart grid is the next generation electric power system, which Supports bidirectional energy
and information flow between Service providers and consumers achieving higher stability,
Reliability, and efficiency. Internet of Things (IoT) technology Is gaining increasing attention in
the recent years. This technology can be used for enabling the smart grid to achieve its Goals
in monitoring, protecting, and controlling through the Incorporation of different metering
devices such as sensors And actuators, while supporting various network functions And
system automation. IoT technology has been applied in Smart buildings, healthcare systems,
agriculture, smart cities, And smart homes, among others application domains. In particular,
the applications of smart home and home energy Management systems (HEMS) are essential
towards achieving energy efficiency. To build such management systems, It is necessary to
identify and control appliances with higher Electrical consumption, i.e., major appliances. The
home Appliances are mostly used for routine household tasks, such As doing laundry, food
preservation, or cooking. Among these Common loads, there are the washing machine, the
heating, Ventilating and air conditioning (HVAC), the dishwasher, The freezer, and the electric
vehicles (EVs). In case of electric vehicles, it is expected that EVs will be a key part Of the future
smart grid as they bring many environmental And economic benefits. However, the
bidirectional operation Flow of EVs carries complex problems into the distribution power
network, which challenge their integration. This bidirectional energy flow is between electric
vehicles and the Power grid. By identifying appliance usage it will be possible to deploy
Multiple useful smart grid applications, such as load planning, demand response (DR) and the
classification of activities Of daily living (ADLs) [7]. The process which enables the recognition
and monitoring the main loads in a household Is commonly known as load monitoring.
There are two Main categories for classifying the methods that manage such Process, software
based methods (non-intrusive load monitoring, NILM) and hardware based methods
(intrusive load Monitoring, ILM) from bottom to up, the task is to recognize Individual
appliance loads through different stages: Data Acquisition, Feature extraction, and
Classification. Stage one Is the process in which data is gathered using a physical Device. Then,
a further processing (Feature extraction) is performed over the acquired samples to obtain a
signature which Represents the appliance electrical consumption. Finally, the Resultant
extraction is frequently classified through machine Learning (ML) models. Both methods share
some common characteristics. One of them is the use of ML techniques for Predicting the
behavior of appliances usage and translating Raw data (e.g., current, voltage and power) into
an easy and Understandable form. These techniques also allow to make A deeper analysis of
the electricity consumption so that it Is possible to build a consumer profile, bringing privacy
Concerns into account. In software-based methods, measurements come from one single
point of sensing (usually the smart meter device). This simplicity for collecting data makes
NILM an attractive solution specially for low-cost implementation scenarios. Even though
these solutions have been the target of Vast majority of research studies in load monitoring
during the last decade, the solutions have shown less accuracy and higher difficulty for its
deployment in practical Scenarios compared to hardware-based methods. The NILM
Techniques are primarily based on event detection. They consist of sampling the aggregated
signal obtained with smart Meters, and then identifying particular signatures of electrical
appliances. The aggregated signal is characterized for Being noisy, which causes that only a
few major appliances Can be classified, depending on the sampling frequency. Even with the
advancement in artificial intelligence (AI), The situation is the same, only some appliances,
e.g., oven, Air-conditioner, washing machine, and EV can be monitored.
Objectives :-
The primary objective of this project is to develop a comprehensive and innovative appliance
recognition module that significantly enhances the efficiency and reliability of Intrusive Load
Monitoring (ILM) systems within smart home environments. As the demand for energy-
efficient solutions continues to rise, particularly with the increasing adoption of Electric
Vehicles (EVs), there is a pressing need for advanced monitoring systems that can accurately
identify and analyze the energy consumption patterns of various household appliances,
including EVs.
This project specifically targets the development of a robust framework that can effectively
monitor and classify the energy consumption associated with different appliances in real-time.
By focusing on the intricacies of appliance behavior, the project aims to create a system that
not only identifies individual appliances but also understands their operational patterns and
interactions. This understanding is crucial for providing users with actionable insights into
their energy usage, enabling them to make informed decisions about their consumption
habits.
To achieve this objective, the project will implement advanced algorithms for data pre-
processing, training, and inference processes. The feature extraction phase will involve the
creation of a ten-feature vector that captures essential statistics behind appliance power
measurements, including minimum, maximum, mean power values, standard deviation,
skewness, kurtosis, coefficient of variation, median absolute deviation, and counts of values
above the mean and in standby mode. These features will serve as the foundation for training
machine learning models that can accurately classify appliances based on their energy
consumption profiles.
Moreover, the project will explore various machine learning techniques to enhance the
classification accuracy of the appliance recognition system. By employing supervised learning
methods, the project aims to train classifiers that can distinguish between different appliances
and their operational states. The integration of multiple datasets, including the UK-DALE
dataset, will provide a diverse range of appliance profiles, further improving the model’s
robustness and generalizability.
Literature Review:-
Sr.No Paper Name Author Description
1. Internet of Things Sanjeevikumar Energy Internet (EI) has been
applications as energy Padmanaban, recently introduced as a new
internet in Ersan Kabalci, concept, which aims to evolve
smart grids and smart Yasin Kabalci smart grids by integrating several
environments. energy forms into an extremely
flexible and effective grid.
2. New Challenges in the Sen Tan, Microgrids (MGs), referred to as
Design of Microgrid Yanpeng Wu, next-generation power systems,
Systems: Communication Peilin Xie are receiving considerable
Networks, Cyberattacks, attention from both industry and
and Resilience academia. Integrated with
distributed energy resources
(DERs), MGs function as a
localized power grid that can be
operated independently or
connected to utility grids.
3. Convergence and Ying Wu, The Energy Internet is proposed
interoperability for the Yanpeng Wu, to enhance the collaborative
energy internet: From Josep M. utilization of distributed
ubiquitous connection to Guerrero renewable energy resources;
distributed automation enable a flexible, customer-
engaged energy transaction
network; and achieve real-time
balancing of supply and demand.
4. A survey on home energy Joaquim Leitão, This work aims to survey the most
management Paulo Gil, recent literature on home energy
Bernardete management systems, providing
Ribeiro an aggregated and unified
perspective in the context of
residential buildings.
5. Opportunities and C. Liu, This paper will reveal the
challenge of vehicle-to- K. T. Chau, opportunities, challenges, and
home, vehicle-to-vehicle, D. Wu. strategies in relation to
and vehicle-to-grid developing EV energy storage.
technologies

Methodology:-
The methodology for this project is structured into several key phases, each designed to
systematically address the objectives of developing an innovative appliance recognition
module for Intrusive Load Monitoring (ILM) systems in smart homes. The following outlines
the detailed steps involved in the methodology:
1.Literature Review and Requirement Analysis. -
Conduct a comprehensive literature review to understand existing ILM techniques, appliance
recognition methods, and the role of Electric Vehicles (EVs) in smart home energy
consumption. -
Identify gaps in current methodologies and establish the specific requirements for the
proposed appliance recognition module, focusing on accuracy, efficiency, and real-time
monitoring capabilities.

2.Data Collection. -
Utilize publicly available datasets, such as the UK-DALE dataset and Dataport residential data,
to gather energy consumption profiles of various appliances, including EVs, ovens, and
microwaves. -
Filter and preprocess the data to create a manageable subset that includes relevant features
for training the recognition module. This subset will focus on three appliances from a specific
household to ensure clarity and relevance.
3. Data Pre-processing. -
Implement data cleaning techniques to handle missing values, outliers, and inconsistencies in
the datasets. -
Perform feature extraction to derive meaningful statistics from the raw data, including
minimum, maximum, mean power values, standard deviation, skewness, kurtosis, coefficient
of variation, median absolute deviation, and counts of values above the mean and in standby
mode. - Normalize
and scale the features to ensure uniformity and improve the performance of machine learning
algorithms.
4. Model Development. -
Select appropriate machine learning algorithms for appliance classification, including
supervised learning techniques such as Support Vector Machines (SVM), Decision Trees, and
Random Forests. -
Split the dataset into training, validation, and test sets to evaluate the performance of the
models effectively. -
Train the models using the training dataset, optimizing hyperparameters to enhance
classification accuracy.
5. Model Evaluation. -
Assess the performance of the trained models using the validation dataset, employing metrics
such as accuracy, precision, recall, F1-score, and confusion matrices to evaluate their
effectiveness. -
Conduct cross-validation to ensure the robustness and generalizability of the models across
different datasets and appliance profiles.
6. Feature Importance Analysis. -
Analyze the importance of each feature in the classification process to identify which
characteristics contribute most significantly to the model’s predictions. -
Utilize techniques such as permutation importance or SHAP (Shapley Additive exPlanations)
values to gain insights into feature relevance.
7. Integration of IoT and Smart Meter Data. -
Develop a framework for integrating data from IoT devices and smart meters to enhance the
appliance recognition system’s capabilities. -
Implement real-time data streaming and processing to allow for continuous monitoring and
feedback on energy consumption patterns.
8. System Implementation and Testing. -
Build a prototype of the appliance recognition module, incorporating the trained models and
the integrated data framework. -
Conduct extensive testing in a controlled environment to evaluate the system’s performance,
reliability, and user interface.
9. User Feedback and Iteration. -
Gather feedback from potential users regarding the usability and effectiveness of the
appliance recognition module. -
Iterate on the design and functionality based on user input, making necessary adjustments to
improve the overall user experience.

10. Documentation and Reporting. -

Document the entire methodology, including data sources, algorithms used, model
performance, and user feedback. -
Prepare a comprehensive report detailing the findings, challenges faced, and
recommendations for future work in the field of ILM and smart home energy management.
By following this structured methodology, the project aims to develop a robust and effective
appliance recognition module that enhances the capabilities of ILM systems, ultimately
contributing to improved energy efficiency and user engagement in smart home
environments.

Expected Outcomes:-
The expected outcomes of this research include:
1. Robust Framework: A well-defined and adaptable framework for IoT-based appliance
recognition that can be implemented in various smart home environments. The framework
will be designed to accommodate different types of appliances and usage patterns, making it
versatile for diverse applications.
2. Enhanced Accuracy: Improved accuracy in appliance classification, leading to better energy
management and insights for users. The framework is expected to provide users with detailed
information about their energy consumption, enabling them to identify energy-saving
opportunities and reduce costs.
3. Contribution to Smart Home Technology: The findings will contribute to the field of smart
home technology by providing a comprehensive solution for load monitoring and appliance
recognition. This research will pave the way for future advancements in energy management
systems, supporting the broader goal of sustainable energy consumption.
4. Integration of Diverse Data Sources: The successful integration of data from IoT devices and
smart meters will create a more holistic view of energy consumption in smart homes. This
outcome will enable the system to leverage real-time data for improved accuracy and
responsiveness, ultimately enhancing the user experience.
5. User-Centric Design and Usability: The project will prioritize user feedback throughout the
development process, resulting in a user-friendly interface for the appliance recognition
module. This outcome will ensure that the system is accessible and easy to use for a wide
range of users, promoting greater adoption of energy monitoring technologies.

6. Contribution to Energy Efficiency and Sustainability: By providing users with detailed

insights into their energy consumption patterns, the project is expected to promote energy
efficiency and sustainability. Users will be empowered to identify high-energy-consuming
appliances and adjust their usage accordingly, leading to reduced energy waste and lower
environmental impact.

Conclusion:-
This study aims to address the challenges associated with appliance recognition in smart
homes through the development of an innovative IoT-based framework. By integrating
advanced machine learning techniques and leveraging real-world datasets, the proposed
solution is expected to significantly enhance energy efficiency and user experience in
residential settings.
The research findings will not only contribute to the ongoing efforts in smart home technology
but also support the broader goal of sustainable energy management, ultimately benefiting
both consumers and the environment. The successful implementation of this framework
could lead to widespread adoption of smart home technologies, fostering a culture of energy
conservation and sustainability in modern households.
In conclusion, this project aims to develop a cutting-edge appliance recognition module that
significantly enhances the capabilities of Intrusive Load Monitoring (ILM) systems in smart
homes, with a particular focus on Electric Vehicles (EVs). By employing advanced machine
learning techniques and integrating diverse data sources, the project seeks to achieve high
accuracy in appliance classification and real-time monitoring, empowering users to make
informed decisions about their energy consumption.
The anticipated outcomes, including improved energy efficiency, user-friendly interfaces, and
comprehensive feature analysis, will not only contribute to individual household savings but
also promote broader sustainability goals. Furthermore, the insights gained from this project
will serve as a foundation for future research and development in the field, fostering
innovation in energy management solutions.
Ultimately, the successful implementation of this project has the potential to transform how
consumers interact with their energy usage, leading to a more sustainable and efficient
approach to energy consumption in smart homes. By bridging the gap between technology
and user engagement, this project aspires to create a lasting impact on energy management
practices, benefiting both users and the environment.

References:-
[1] Y. Kabalci, E. Kabalci, S. Padmanaban, J. B. Holm-Nielsen, and F. Blaabjerg, ‘‘Internet of
Things applications as energy internet in Smart grids and smart environments,’’ Electronics,
vol. 8, no. 9, p. 972, Aug. 2019.

[2] S. Tan, Y. Wu, P. Xie, J. M. Guerrero, J. C. Vasquez, and A. Abusorrah, ‘‘New challenges in
the design of microgrid systems: Communication Networks, cyberattacks, and resilience,’’ IEEE
Electrific. Mag., vol. 8, No. 4, pp. 98–106, Dec. 2020.
[3] Y. Wu, Y. Wu, J. M. Guerrero, J. C. Vasquez, E. J. Palacios-Garcia, and J. Li, ‘‘Convergence and
interoperability for the energy internet: From Ubiquitous connection to distributed
automation,’’ IEEE Ind. Electron. Mag., vol. 14, no. 4, pp. 91–105, Dec. 2020.
[4] J. Leitao, P. Gil, B. Ribeiro, and A. Cardoso, ‘‘A survey on home energy Management,’’ IEEE
Access, vol. 8, pp. 5699–5722, 2020.
[5] B. Hartono, S. P. Mursid, and S. Prajogo, ‘‘Review: Home energy management system in a
smart grid scheme to improve reliability of power Systems,’’ IOP Conf. Ser., Earth Environ. Sci.,
vol. 105, Jan. 2018, Art. No. 012081.
[6] C. Liu, K. T. Chau, D. Wu, and S. Gao, ‘‘Opportunities and challenges Of vehicle-to-home,
vehicle-to-vehicle, and vehicle-to-grid technologies,’’ Proc. IEEE, vol. 101, no. 11, pp. 2409–
2427, Nov. 2013.
[7] M. A. Devlin and B. P. Hayes, ‘‘Non-intrusive load monitoring and classification of activities
of daily living using residential smart meter data,’’ IEEE Trans. Consum. Electron., vol. 65, no.
3, pp. 339–348, Aug. 2019.

[8] X. Yuan, P. Han, Y. Duan, R. E. Alden, V. Rallabandi, and D. M. Ionel, ‘‘Residential electrical
load monitoring and modeling–state of the art and Future trends for smart homes and grids,’’
Electr. Power Compon. Syst., Vol. 48, pp. 1–19, Nov. 2020.
[9] D. Burbano, ‘‘Intrusive and non-intrusive load monitoring (a survey) inference and learning
approach,’’ Latin Amer. J. Comput., vol. 2, no. 1, p. 9, 2015.
[10] A. U. Rehman, T. T. Lie, B. Vallès, and S. Tito, ‘‘Low complexity event Detection algorithm
for non-intrusive load monitoring systems,’’ in Proc. IEEE Innov. Smart Grid Technol.-Asia (ISGT
Asia), May 2018, p. 751.

[11] I. J. Goodfellow, Y. Bengio, and A. Courville, Deep Learning. Cambridge, MA, USA: MIT
Press, 2016.
[12] UK Domestic Appliance-Level Electricity (UK-DALE) Dataset. Accessed: Feb. 2021. [Online].
Available: https://jack-kelly.com/data/

[13] Dataport | Residential Data. Accessed: Apr. 2021. [Online]. Available:

https://dataport.pecanstreet.org/
[14] Electric Cars, Solar & Clean Energy. Accessed: May 2021. [Online]. Available:
https://www.tesla.com/

[15] R. Shanmugam and R. Chattamvelli, Statistics for Scientists and Engineers. Hoboken, NJ,
USA: Wiley, 2015.