The Internet of Trains

Frank Saeuberlich (Teradata)
Location: 127-128
Slides:   1-PPTX 

How can big data make your journey to work better? In this case study we’ll explore how! Trains today are complex systems consisting of many embedded subsystems, which operate together with the overall goal of delivering a high quality transportation service. The volume of data generated through logging of diagnostic events and sensor readings is huge and can be an untapped source with potential for new business insights. There are many issues involved in how to handle these data volumes and their underlying data types such as XML, JSON, binary data and custom logs. Then the data needs to be explored for relevant information and the data pre-processing becomes important, i.e. aggregating sensor readings to meaningful time intervals. What analytics can be performed on this data and what value is hidden in the patterns revealed by effectively storing and parsing this data? We will explore these opportunities and discuss different approaches with the objective of reducing the cost of rolling stock fleet maintenance, minimizing train downtime and improving fleet reliability.

Photo of Frank Saeuberlich

Frank Saeuberlich


Dr. Frank Saeuberlich is Director Advanced Analytics as part of the Teradata International Data Science team. He previously worked at Urban Science International where he was responsible for Customer Analytics. In this role he has worked with client teams to implement analytical solutions and he pioneered new types of analysis to improve the efficiency of automotive clients’ marketing efforts. Prior to that as European Customer Solutions Practice Manager he was responsible for the Urban Science Customer Solutions practice in Europe.
While working on his doctoral thesis, he was a scientific assistant at the Institute of Decision Theory and Management Science, University of Karlsruhe, Germany where he developed a course in web mining and supervised several research papers.