Virtual Field Service Ecosystem (VSE) using AR (Augmented Reality) collaboration with SiemensAG
Master thesis
View/ Open
Date
2018-06-15Metadata
Show full item recordCollections
- Studentoppgaver (TN-IDE) [901]
Abstract
With the huge advancement of technologies, our viewpoint to see, hear, observe and
feel the surroundings around us is changing every single moment. Building a virtual
ecosystem is an idea which needs much time and effort. The purpose of this project is
building an ecosystem with AR applications combining machine learning features. In
this way, users can gain access to information in a very interactive, contextualized ways
which provide a deeper understanding of the physical problems around them and how
to solve them easily. A smart machine learning algorithm is only possible if the data
provided is concrete and huge to perform any thorough analysis. In this experiment, huge
data containing different significant features of a single feed production machine from
Siemens is provided where the quality of the product is depended on pressure. Analysis
of that data is performed showing graphs, selecting features, validations, mathematical
implementations or statistical analysis to propose a model. The significant part of the
model building is predicting pressure value for advanced maintenance of the machine
and accuracy of the model must be high. The predicted data, analysis of graphs and
validation results is proposed to be stored on a cloud system. The AR application is
supposed to show ML results. This includes showing every data that is stored in the
cloud in the AR application. That way the AR and machine learning are combined in a
single application which has the possibility to be extended later for bigger solutions.
Description
Master's thesis in Computer science