The rapid advancements in smartphone sensor technology over the past decade have led to significant improvements in accuracy, power efficiency, and cost, enabling their widespread adoption in various domains. One of the most impactful applications of inbuilt smartphone sensors is Human Activity Recognition (HAR), which involves identifying activities of daily living (ADLs) using sensor data and machine learning models. This chapter provides a foundational understanding of HAR, discussing its classification based on sensor modalities, including wearable, extrinsic, vision-based, and hybrid approaches. It highlights the significance of HAR in industries such as healthcare, sports, security, and well-being, where real-time activity monitoring plays a crucial role. Despite its potential, HAR systems face challenges related to data acquisition in uncontrolled environments, sensor calibration, optimal feature selection, and model efficiency. The chapter also outlines the motivation behind developing a robust smartphone sensor-based HAR system and the need for an optimized framework that balances computational efficiency, real-time performance, and accuracy. Finally, the major contributions of this book are summarized, including dataset development, feature selection techniques, and the design of a variable deep learning framework. The chapter concludes with an overview of the book’s organization, detailing the key topics covered in subsequent chapters.