Visual Analytics Methods to Explore Drilling Data: Pattern Recognition and Trend Analysis in Multi-variate Drilling Data

Problem: 

Drilling rigsToday's drilling activities employ a variety of technologies to make the drilling operation as efficient as possible. The company TDE provides solutions to improve drilling performance and reduce well construction cost. The basis to achieve this objective is the ability to manage and analyze drilling data in a unique way. These data are collected using various sensors attached to the drilling rigs. TDE applies a variety of methods to extract key features about these data, and optimize the overall drilling performance.

The data analysis of drilling data is challenging due to the large volume of data collected over weeks (in 1-10 second interval) and due to the existence of outliers and missing values.

The field of Visual Analytics aims to combine advantage of both visual and analytical methods and to utilize results from cognition and perception sciences to provide the analyst with more powerful analysis tools.

Aim: 

This work aims to develop Visual Analytics methods for better understanding both the drilling data collected by the sensors and the extracted features. These methods should also help in discovering non-trivial patterns and trends in the data.
The use of visual methods gives more potential for understanding the data, what values they actually take, and what correlations do exist. Also, outliers and missing values can be better detected. The use of automatic (data mining) methods gives more potential for analyzing large data, and for performing elaborate computations with these data.

Topics: 
Visual Analytics, Data Mining, Information Visualization, Time Series Analysis
Other information: 

Conditions:

  • TDE will provide large volumes of data to be used for developing prototypes and testing them.
  • TDE will provide 1-2 weeks in-house schooling, so that you gain detailed understanding of the data, the tasks, and the users.
  • CVAST will provide scientific supervision and assistance.
  • Close cooperation between TDE and CVAST to progress the research and to exchange results.
  • The work will be partially financed from TDE.

Literature:

Duration: 6 Months

Sponsor: TDE Thonhauser Data Engineering GmbH

Previous knowledge: 
previous experience with data analysis (in projects or exercises)
Scope: 
MA
Contact: 
Silvia Miksch, by appointment, miksch [at] ifs.tuwien.ac.at
Area: 
Visual Analytics (VA)
Status: 
open