Coughing is the sudden expulsion of air from the lungs to clear the breathing passageways of unwanted irritants. However, if it sounds too loud and occurs for long periods, then it will cause a person chest pain, difficulty breathing and a high fever which requires treatment from a doctor. Knowing the seriousness of the matter, this research develops a cough detection system using machine learning technique. The system employed five distinct modules namely audio sampling, sound feature extraction, model training, cough detection, and audio data testing. Via the modules, when a person coughs then an important audio signal characteristic will be retrieved. It became an input for an Artificial Neural Network (ANN) model that has been trained with a cough sound dataset. The model will analyze the sound for a coughing fit and identify whether the sound is a cough or vice versa. A real-time test was arranged to test the performance of the model. It was programmed in an embedded controller and then installed to a device namely a Modular and Open System (MOST). The results demonstrate the model has a precision of 97.5% and successfully detects cough from a user.