Important Things to Consider When Implementing IIoT, Advanced Analytics, and Big Data
CIOREVIEW >> Energy >>

Important Things to Consider When Implementing IIoT, Advanced Analytics, and Big Data

Craig Harclerode, Global O&G Business Development Executive, OSIsoft
Craig Harclerode, Global O&G Business Development Executive, OSIsoft

Craig Harclerode, Global O&G Business Development Executive, OSIsoft

We are all inundated by IIOT, advanced analytics, and “Big Data” claims of transformative business value. Start now, start big, and move fast. This marketing pressure can be confusing, and result in misfires, disillusionment, and doomed initiatives that do not deliver on business value, consume scarce capital and, even more importantly, soak up organizational bandwidth with lost opportunity costs.

Industrial and commercial customers, however, should move cautiously. The significant challenges include cyber security and data governance; gathering, normalizing and integration of the IIoT “lots of little pipes” operational data with existing industrial data Operations Technology (OT) fabric from “big pipe” data sources like SCADA and DCS.

These are key considerations because companies are not just collecting industrial data to analyze at a future date, like ad companies trying to micro-target demographic segments. Their entire operation depends on real-time visibility and the ability to understand and control their processes safely and optimally in real time. Without rock-solid reliability, highly secure, and real-time visibility, productivity, financial viability—even the safety and health of their workers and communities where they operate—can be imperiled. A more conservative approach is required.

Do not take this as a negative towards moving forward with IIoT and advanced analytics. There are many successful use cases that reinforce the opportunity.

For example, MOL PLC, a Hungarian inte­grated O&G company, has presented in sev­eral public venues the generation of over $500 Mn EBITDA by using advanced analytics for things like hydrogen em­brittlement corrosion and the application of machine learning in several refining processes. Oth­er areas include dynamic Integrity Operating Windows (IOW), and advanced predictive CBM. (Ref­erence OSIsoft 2016 UC)

It is more about how versus if. MOL, like other companies, used a very prescriptive approach to their IIoT, advanced analytics, and big data journey.

My Perspective on Doing Things Differently:

1. They did not forget that it is about delivering business value, supporting a business strategy, and achieving a return on investment, not applying IIoT and advanced analytics for technology sake. Start small and strategically with a sound business use case, end-user input, support, and joint IT/OT accountability.

2. They started the journey of creating an operational data infrastructure (OT) as a foundational element of an overarching IIoT, advanced analytics, and big data strategy. If you listen to the marketing hype, you may have come to the conclusion that all of your data will end up in a data lake in the cloud, one massive (and growing) storehouse that will contain everything from sensor data to customer records. Successful implementations leverage fit for purpose technologies to address the unique characteristics and challenges of time series data.

 Successful implementations leverage fit for purpose technologies to address the unique characteristics and challenges of time series data and real time analytics   

This OT data infrastructure includes: the access of associated OT metadata to enable an “OT chart of accounts” where all OT data can be aggregated across a portfolio of OT data sources including on premise and cloud based IIoT sources; abstraction and normalization of tagging, asset names, units of measure, and time zones; quality assurance; high fidelity time series archival; and contextual organization analogous to the financial or “IT chart of accounts”, which structure has been mandated by regulations. Lastly, the OT data infrastructure needs to be a hybrid architecture with an integrated on-premise and cloud architecture to enable flexibility and evolution over time as technology and companies continue to change.

3. Within the OT data infrastructure, they created configurable smart asset model templates that can be leveraged to standardize the structure and analytics of asset classes enabling roll out at scale and pace across an enterprise. These smart asset models for assets such as heat exchangers, and pumps are used to perform affront line analytics such as efficiency, run times, CBM in effect acting as analytics preprocessors to higher level advanced analytics and big data. As these analytics move to the edge, the smart object models can hybridize to cover and support all three layers: at the edge, in the OT data infrastructure, and support of higher-level analytics. Shell reported having over 500 smart asset templates at a recent user’s conference.

4. They rationalized what and where analytics are performed between the OT data infrastructure, advanced analytical platforms, and “big data” with as much OT analytics done in the infrastructure via smart models providing in effect an analytical preprocessor to higher-level analytics. Calculations such as exchanger and pump efficiencies, energy utilization, and yields or advanced CBM can and should be done in the OT infrastructure closer to the assets vs the higher-level platforms and the results uploaded for many reasons ranging from efficiency, ability to leverage across multiple end uses, and ability to operationalize. Performing OT analytics in the OT data infrastructure will also enable the migration of analytics to the edge over time.

5. They bridged OT-IT by use of the OT data model to bring the IT and OT data systems together. This can be accelerated by the use of an integration layer that cleans, augments, shapes and transmits (CASTs) operational data so that it can be consumed in unstructured IT systems. These data integrators effectively automate data preparation and translation.

As example, Cemex, one of the world’s largest cement manufacturers has adopted integrator technologies to deliver data from its 70 plants to its business intelligence applications for preparing reports. By Cemex’s estimates, the time to extract data from 70 sites for production reports has declined from 744 hours to 5 minutes while data preparation was cut from 3 days to less than a minute.

IIoT, advanced analytics, and big data are here and growing, make no mistake. They will dramatically transform our largest and oldest industries. If you approach their implementation and use strategically with an approach presented above, you will increase the probability of sustainable value from your IIoT, advanced analytics, and big data initiatives.

Read Also

Digital Transformation & Innovation

Carlos Andre Sant'Anna, Chief Digital Officer, JHSF

Digital Transformation and technological advancements in a NEO Bank

Matthias Fengler, Head of Finance & Controlling, N26

Digitising your businesses DNA

Fraser Collins, Group Head of Commercial Finance, International Personal Finance (IPF)

The Bank's Experience: How a Company's Use of Fintech Can Accelerate...

Mārtiņš Bērziņš, Head of Digital Customer Experience, Deputy Business Development, Citadele Bank

Fintech solutions for the exploding savings market: How banks can...

Paul Knodel, CEO and Managing Director, Raisin US Inc.

Looking to Finance a Tech Startup? Your Timing May Be Just Right

Kurt Nichols, Managing Director, Portfolio Manager, CIBC Innovation Banking