This paper discusses the development and application of data-driven Soft Sensors in the process industry. Soft Sensors are predictive models that use historical data to monitor and control processes, particularly in industries such as chemical, bioprocess, and steel. The authors highlight the characteristics of process industry data, which include missing values, outliers, drifting data, co-linearity, and varying sampling rates. They emphasize the importance of data preprocessing techniques, such as handling missing values, outlier detection, and feature selection, to improve model performance. The paper also reviews various modeling techniques, including Principal Component Analysis (PCA), Partial Least Squares (PLS), Artificial Neural Networks (ANN), Neuro-Fuzzy Systems, and Support Vector Machines (SVM). It provides a comprehensive overview of Soft Sensor applications, focusing on online prediction, process monitoring, and fault detection. The authors discuss the challenges and solutions in Soft Sensor development and maintenance, emphasizing the need for automated adaptation mechanisms to address changing data conditions. The paper concludes with a discussion on future research directions in the field of Soft Sensors.This paper discusses the development and application of data-driven Soft Sensors in the process industry. Soft Sensors are predictive models that use historical data to monitor and control processes, particularly in industries such as chemical, bioprocess, and steel. The authors highlight the characteristics of process industry data, which include missing values, outliers, drifting data, co-linearity, and varying sampling rates. They emphasize the importance of data preprocessing techniques, such as handling missing values, outlier detection, and feature selection, to improve model performance. The paper also reviews various modeling techniques, including Principal Component Analysis (PCA), Partial Least Squares (PLS), Artificial Neural Networks (ANN), Neuro-Fuzzy Systems, and Support Vector Machines (SVM). It provides a comprehensive overview of Soft Sensor applications, focusing on online prediction, process monitoring, and fault detection. The authors discuss the challenges and solutions in Soft Sensor development and maintenance, emphasizing the need for automated adaptation mechanisms to address changing data conditions. The paper concludes with a discussion on future research directions in the field of Soft Sensors.