Intelligent supply and demand for marine protein factory (based on MindSphere platform)
MetadataShow full item record
- Studentoppgaver (TN-IDE) 
The purpose of this project is to collect Internet Of Things data from available sources in the value chain. Our viewpoint to observe IoT data streams from all different processes in the factory as well as get an understanding of the different available features and the relation between them, by learning about the process with Siemens engineers and the factory. According to the captured data, the motivation is to model and predict production efficiency by using some appropriate machine learning algorithm. The models provide insights and correlations between features from the input parameters, and also analyze the data to produce the output. Data Pre-processing and re-sampling techniques are necessary to provide a deeper understanding of the essential features and to know which parameters are highly influential on production efficiency. In this experiment, captured data contains the different essential features of fish oil and meal production machines that are provided by Siemens. The significant part of the project is to analyze the data that has performed selecting features, validations, statistical analysis as well as presenting some graphs to decide which model can provide the best prediction with less error and finally propose a model for prediction of one of the critical key performance indicator.
Master's thesis in Computer science