GO Digital Energy Oil and Gas Middle East

McKinsey survey finds digital initiatives floundering. Petroleum Development Oman ARV project improves SAP data quality with machine learning. AVEVA befriends the data monster.

Anders Brun presented the results of a McKinsey survey of 1800 execs working in asset intensive industries. Strong support for ‘digital and analytics’ (DnA) as 91% believe such will ‘materially change their industry’. Brun offered some choice quotes from major oils’ C-suites on the expected benefits of AA. Across the O&G value chain, McKinsey sees a ‘transformational agenda’ emerging centered around 15 main themes. However, the McKinsey survey found that overall, only 17% were actually sponsoring large digital initiatives and as few as 2% reported seeing ‘material and sustainable benefits from DnA’. So what is going wrong?

Brun cites a number of failings. A mindset that focuses on technology rather than business impact. The difficulty of navigating the technology jungle where ‘everybody offers everything’. Organizational governance is not fit for purpose and there is a general lack of usable data, many firms are struggling to deliver value due to data challenges. 72% noted managing data as being one of the top challenges to scaling data and analytics impact.

In an inditement of 20 plus years of data management initiatives, McKinsey found a lack of buy-in from business leaders and data governance that exists only ‘on paper’ as opposed to being actioned. Data architecture development requires much more time and investment than realized and data itself is too often perceived as ‘IT-stuff’ as opposed to a business asset. Finally, there is a lack of data talent such as data architects, data engineers and data visualization experts.

Oil and gas companies are keen on executing digital pilots and proofs of concept but fail to go the ‘last-mile’ and realize the actual benefits. What is needed has been formulated as the McKinsey DnA Tech-Enablement Playbook for accelerating digital transformation. This includes findings from the top quartile digital performers who for instance, spend over 50% of their budget on changes to the ‘last mile’ of business processes. Other recommendations include an operating model designed around collaborative, cross-functional teams and clear accountability and decision-making pushed down to the working team-level. Download the Playbook for Utilities.

Aleksandr Zykov showed how Petroleum Development Oman has used machine learning to improve data quality in an asset register verification (ARV) project. Incomplete and erroneous values in PDO’s SAP Asset Register were a contributing cause of process safety incidents. The ARV project set out to align SAP technical objects with process engineering flow schemas and P&IDs. Attempts to do this manually were time consuming and error prone. PDO developed an ARV process improvement toolkit that learns from previous object matches and provides quality control of incoming technical data.

The model uses regular expression matching between SAP functional location and P&ID tags right along the safety critical systems pipeline. An analysis of weighted matches determines which of various possible values is statistically most likely to be correct. Object matching time is now down to 0.09 sec/object. The whole 180,000 set of process containment equipment tags can be QC’d in one pass. PDO is now looking to apply the toolkit to other data matching exercises such as real time operations PI-to-SAP plant maintenance data. The process has identified hitherto undocumented functional locations. These have been fixed with ‘focused data harvesting’.

Interestingly, in the light of the McKinsey presentation above, Amit Kar (Aveva) thinks that the ‘data monster’ is your friend? Once you acknowledge this, like a good friend, the monster will tell you the truth about yourself. But this is a relationship that needs to stay fresh! Kar, citing a different McKinsey survey, forecasted that operations would be ‘fully infused’ with AI in the next few years with a ‘+122% impact on cashflow’. Companies making AI progress by 2030 will see a +10% impact and the AI laggards will get -23%, due to eroding competitiveness. In all events, Aveva is there to capture plant data in its ‘living’ digital twin with added cloud data storage and AI capabilities. Although Kar did not cover this in his talk, the Aveva’s acquisition of OSIsoft, the developer of the ubiquitous PI System has hiked Aveva’s data street cred somewhat.

More from GO Digital.

Click here to comment on this article

Click here to view this article in context on a desktop

© Oil IT Journal - all rights reserved.