This paper provides a comprehensive review of 1D Convolutional Neural Networks (1D CNNs) and their applications. 1D CNNs, which are specialized for processing 1D signals, have gained significant attention due to their ability to achieve state-of-the-art performance in various engineering applications, such as personalized biomedical data classification, structural health monitoring, anomaly detection, and motor fault detection. Unlike 2D CNNs, 1D CNNs are more computationally efficient, real-time, and suitable for low-cost hardware implementations. The paper highlights the advantages of 1D CNNs, including their ability to combine feature extraction and classification tasks into a single model, and their adaptability to limited labeled data and high signal variations. It also discusses the challenges and limitations of 1D CNNs, such as computational complexity and the need for specialized hardware. The paper covers the general architecture and principles of 1D CNNs, their training methods, and their applications in detail. Additionally, it presents benchmark datasets and software tools used in these applications, which are publicly shared. The paper concludes by discussing future directions for research in 1D CNNs.This paper provides a comprehensive review of 1D Convolutional Neural Networks (1D CNNs) and their applications. 1D CNNs, which are specialized for processing 1D signals, have gained significant attention due to their ability to achieve state-of-the-art performance in various engineering applications, such as personalized biomedical data classification, structural health monitoring, anomaly detection, and motor fault detection. Unlike 2D CNNs, 1D CNNs are more computationally efficient, real-time, and suitable for low-cost hardware implementations. The paper highlights the advantages of 1D CNNs, including their ability to combine feature extraction and classification tasks into a single model, and their adaptability to limited labeled data and high signal variations. It also discusses the challenges and limitations of 1D CNNs, such as computational complexity and the need for specialized hardware. The paper covers the general architecture and principles of 1D CNNs, their training methods, and their applications in detail. Additionally, it presents benchmark datasets and software tools used in these applications, which are publicly shared. The paper concludes by discussing future directions for research in 1D CNNs.