Our Analytics services involve the use of machine learning algorithms to assimilate Inline Inspection Data and Integrated Direct Assessment Data for developing Predictive Models for Pipeline Operations and Maintenance (O & M) Management. In our experience, we notice that most pipeline owners and operators have in their possession a huge amount of pipeline operations data from the day they conceptualise their pipeline. This design, construction, testing, commissioning, day to day operations and maintenance data are usually archived in various formats including MS Excel, MS Word, Autocad and others. Over the years, some of these data get misplaced and others kept in little known storage spaces and are seldom referenced to the present day recorded O & M Data. Majority of the owners and operators tend to adopt a scheduled proactive maintenance program and others to address their requirements through reactive programs after a failure.
In pipeline operations, safety and reliability are the primary cornerstones of the industry. By harnessing the already available existing data using machine language which is a subset of artificial intelligence, we should be able to visualise various patterns of the various pipeline operations and maintenance parameters to develop a predictive model which provide the additional impetus to optimise the O & M OPEX. These applications can be developed for individual components like valves, flanges, meters and others to the full length of the pipeline using soil parameters, ROW inspection records, CP system monitoring data, ILI inspection data, NDT data, corrosion monitoring probe data, coating repair testing programs etc. etc. Overall, the challenge is defining the pertinent data and letting the algorithm deliver the various possible scenarios for sound O & M decision making.