Dataset visualisation, analysis and interpretation using Artificial Intelligence

We developed the Quality Distributed Optical Sensing (Q-DOS) software platform for our client Sensalytx for use by a confidential major international energy company to visualise and analyse a large oil & gas well Distributed Optical Sensing (DOS) data set. The Q-DOS software platform helped to characterise production efficiency and the changing down-hole conditions by visualising and analysing historical and up to date data.

The Q-DOS digital platform enables visualisation, analysis, and interpretation of DOS data, which addresses both Distributed Temperature Sensing (DTS) and Distributed Acoustic Sensing (DAS) values. Thermal and acoustic readings are acquired along the whole length of the fibre optic cables, often over several kilometres, which behave as continuous sequences of sensors in a distributed monitoring network. Q-DOS allowed correlation with other well-relevant datasets such as pressure, lithology, gamma rays, completion and inclination.

The fundamental approach of Q-DOS is applicable to optical fibre sensing applications in many other industries and in a variety of different contexts.

Automated dataset analysis using Artificial Intelligence

Industries are facing challenges in the handling of ever-increasing quantities of data; the upstream Oil and Gas sector in particular is severely affected by this “data deluge” challenge and is at risk of failing to extract valuable information from their exploration and production data.

So far, numerical data cleaning and curation have been considered as synonyms for numerical data quality improvement, but the latter is much more than simply addressing gaps, outliers, noise and bias in numerical datasets typically composed of time series of some kind.

HyperDap has invested in its Intelligent Data Quality Improver (IDQI) tool to address the issue of digital exploration and production data quality in the Oil and Gas industry, but it is also very capable of addressing similar challenges in many other sectors. IDQI built on novel data quality improvement knowledge and technologies previously devised by HyperDAP and drew on guidance and expertise from academic research staff from the University of Aberdeen.

The IDQI project was sponsored by the Oil & Gas Innovation Centre and The Data Lab.

digitalisation assessment

The prospective client was a consulting service company delivering personalised advisory packages to their customers. Their operating model was based on the manual identification of their customer needs and on the provision of customised advice via email or phone. This time- and resource-consuming approach prevented the client from acquiring a much wider market share. The client was aware of the potential benefit of using AI to automate creation and delivery of its advisory packages.

We identified the critical bottleneck in the automated creation of personalised solutions. Working closely with the client assessing the requirements and feasibility of each solution proposed, we identified the AI technologies and solutions that would have allowed automated modelling and reasoning with the required domain knowledge.

We proposed an articulated platform composed of a mobile client for the delivery of personalised, advisory packages and a cloud-based server for all the AI-based processing and the storage of the domain knowledge to be used to generate the personalised advice.

The assessment was conducted on time and on budget. The project was then agreed.