1. To be Published in CIO Review – September, 2016
Important Things to ConsiderWhenImplementingIIoT,AdvancedAnalytics, and Big Data
By Craig Harclerode,Global O&GBusinessDevelopmentExecutive,OSIsoft
Quote:“Successful implementationsleverage fitforpurpose technologiestoaddressthe unique
characteristicsandchallengesof time seriesdata andreal time analytics”
We are all inundatedbyIIOT,advanced analytics,and“BigData”claimsof transformative business
value.Startnow,start big,and move fast. Thismarketingpressure canbe confusing,andresultin
misfires,disillusionment,anddoomedinitiatives thatdon’tdeliveronbusinessvalue,consumescarce
capital and,evenmore importantly,soakup organizationalbandwidth withlostopportunitycosts.
Industrial andcommercial customers,however,shouldmove cautiously. The significantchallenges
include cybersecurityanddatagovernance;gathering,normalizingand integration of the IIOT“lotsof
little pipes”operational datawith existingindustrialdata(OT) fabricfrom“bigpipe”data sourceslike
SCADA and DCS.
These are keyconsiderationsbecausecompaniesaren’tjustcollectingindustrial datatoanalyze ata
future date,like adcompaniestryingtomicro-targetdemographicsegments.Theirentire operation
dependsonreal-timevisibility andthe abilitytounderstandandcontrol theirprocesses safelyand
optimallyinreal time. Withoutrock-solidreliability,highlysecure, andreal-time visibility,productivity,
financial viability—eventhe safetyandhealthof theirworkers andcommunitieswhere theyoperate --
can be imperiled.A more conservative approachis required.
Don’ttake this as a negative towardsmovingforwardwithIIOT andadvancedanalytics.There are many
verysuccessful use casesthatreinforce the opportunity.
For example, MOLPLC,a HungarianintegratedO&Gcompany,has presented inseveral publicvenues
the generationof over$500M EBITDA byusingadvancedanalytics forthingslike hydrogen
embrittlementcorrosion andthe applicationof machine learninginseveral refiningprocesses. Other
areas include dynamicintegrityoperatingwindows(IIOW),andadvancedpredictive CBM.(Reference
OSIsoft2016 UC)
It ismore abouthowversus if.MOL, like othercompanies,usedaveryprescriptive approachtotheir
IIOT,advancedanalytics,andbigdata journey.
What didMOL and otherleadingcompanies dodifferently thatisseparatingthemfromtheirpeers?
From myperspective:
1. Theydid not forgetthat it isabout deliveringbusinessvalue,supportingabusinessstrategy,
and achievinga return on investment,notapplyingIIOTandadvancedanalyticsfortechnology
sake.Start small andstrategically withasoundbusinessuse case,end-userinput,support,and
jointIT/OTaccountability.
2. They started the journeyof creatingan operational data Infrastructure (OT) as a foundational
elementof an overarching IIOT, advanced analytics, and big data strategy. If youlistentothe
marketinghype,youmayhave come to the conclusionthat all of your data will endupin a data
lake inthe cloud,one massive (andgrowing) storehousethatwill containeverythingfrom
2. To be Published in CIO Review – September, 2016
sensordata to customerrecords. Successful implementationsleverage fitforpurpose
technologies toaddressthe unique characteristicsandchallengesof time series data.
ThisOT data infrastructure includes:the accessof associatedOTmetadatato enable an“OT
chart of accounts” where all OTdata can be aggregated acrossa portfolioof OTdata sources
includingonpremandcloudbasedIIOTsources; abstractionand normalizationof tagging,asset
names,units of measure,andtime zones;quality assurance;highfidelitytime series archival;
and contextual organization analogoustothe financial or“IT chart of accounts” whichstructure
has beenmandatedbyregulations. Lastly,the OTdatainfrastructure needstobe a hybrid
architecture with anintegratedon premandcloudarchitecture to enable flexibilityand
evolutionovertime astechnologyandcompaniescontinue tochange.
3. Withinthe OT data infrastructure, they createda configurable smart asset model templates
that can be leveragedtostandardize the structure andanalyticsof assetclasses enablingroll out
at scale and pace acrossan enterprise.These smart assetmodelsforassetssuchas heat
exchangers,andpumpsare usedto performaffrontline analyticssuchasefficiency,runttimes,
CBM ineffectactingas analyticspreprocessorstohigherlevel advancedanalyticsandbigdata.
As these analyticsmove tothe edge,the smartobjectmodelscanhybridizetocoverand
supportall three layers:atthe edge,inthe OT data infrastructure,andsupportof higher-level
analytics.Shell reportedhavingover500 smart assettemplatesata recent user’sconference.
4. They rationalizedwhat and where analytics are performedbetweenthe OT data
infrastructure,advanced analytical platforms, and “bigdata” withas much OT analyticsdone in
the infrastructure viasmartmodels providingineffect ananalytical preprocessortohigherlevel
analytics.Calculationssuchasexchanger andpumpefficiencies,energyutilization,and yieldsor
advancedCBMcan andshouldbe done inthe OT infrastructure closertothe assets vsthe higher
level platformsandthe resultsuploadedformanyreasonsrangingfromefficiency,abilityto
leverage acrossmultiple enduses, andabilitytooperationalize.PerformingOTanalyticsin the
OT data infrastructure will alsoenable the migrationof analyticstothe edge overtime.
5. They bridgedOT-IT by use ofthe OT data model to bring the IT and OT data systems together.
Thiscan be acceleratedbythe use of an integrationlayerthatcleans,augments,shapesand
transmits(CASTs) operational datasothat itcan be consumedinunstructured ITsystems.These
data integratorseffectivelyautomate datapreparationandtranslation.
As example, Cemex,one of the world’slargestcement manufacturers hasadoptedintegrator
technologiestodeliverdatafromits70 plantsto itsbusinessintelligence applicationsfor
preparingreports. ByCemex’sestimates,the time toextractdatafrom70 sitesforproduction
reportshas declinedfrom744 hoursto 5 minutes whiledatapreparation wascutfrom3 daysto
lessthana minute.
IIoT, advanced analytics,andbigdata are here andgrowing,make nomistake. They will dramatically
transformour largestandoldestindustries. If youapproach theirimplementationanduse strategically
withan approach presentedabove,youwillincreasethe probabilityof sustainable valuefromyourIIOT,
advanced analytics,andbigdata initiatives.