MechChem Africa September 2017

⎪ Petrochemical industries, oil and gas ⎪

Louise Steenekamp, southernAfrica director of energy and natural resources for Wipro Limited, speaks about the role of Artificial Intelligence (AI) in the oil & gas sector, outlining the benefits it will offer in navigating an uncertain future characterised by fewer sources of growth, slimmer margins, and increasing complexity. AI in the oil and gas sector

A s oil prices fell from their highs of well over US$100 per barrel, to less than $35 by early 2016, energy and resource companies were forced into a dramatic cost-cutting mode – streamlining focus and creating new efficiencies within their operations. Now, evenwith the resurgence inprice, these industrial firms are realising the benefits of pre- serving operational efficiencies, as they look to scalemore smartly–andmoredigitally– into the future. ByapplyingArtificial Intelligence (AI) and CognitiveLearningtools,industrialfirmswillfind new market opportunities and unlock further productivity gains. By analysing vast swathes of data – from macroeconomic trends, to weather patterns, to consumer spending patterns – energy compa- nies can predict future demand and dynamically resource their operations to capture the most profitable opportunities. By analysing seismic vibrations, reservoir pressure differentials, strata permeability and other geospatial data, AI can effectively guide decisions about where to drill. For instance, Intel’s 2016 acquisition of Nervana Systems, a start-up applying AI to enhance operational ef- ficiency inoil exploration, caught theattentionof many industryplayers, andbecamean important symbol for the future of exploration. Dynamic operations involve embedded sensors that enable better visibility across the value-chain – alerting managers when indus- trial equipment is liable to overheat, rerouting truck schedules to another depot based on traffic conditions, and the like. Many oil and gas players are not yet gathering the richness and depth of data that may be possible as sensor manufacturers jostle with technology firms and oil industry players to commercialise the data they are generating. In terms of storage and transportation of oil and gas, the composition of oil or gas resources can be accurately monitored – temperature, moisture levels, transport time duration, etc. With the right cognitive tools, one can better understand the effect of certain external con- ditions on the composition and profitability of the resources. Oil & gas development players have a golden opportunity tocreate their owndigital platforms or ecosystems, enabling others to create apps and connect to their operations via APIs. With

A I unde r p i nn i ng the development of

these platforms, industrial firms can reach new markets and deepen their expertise by pulling in the specialised services of partners, vendors, and others within the value chain. Due to their geographically dispersed, engineering-intensive nature, the oil & gas industry has a heavy requirement for skilled management and technical staff. As the younger (millennial) workforce takes a fresh view on how they can add value to their company, they need the infrastructure and tools to turn their ambitions to reality. If the average millennial employee is used to AI-powered social media algorithms and digital navigation systems, then the corporate tools must have a similar level of sophistication. It is also important tonote that oil &gas com- panies operate within complex supply chains, so rule-based anomaly detection can be used to detect things like fraud, duplicate payments, untapped volume discounts and other areas of direct cost benefit in the supply chain areas. AI can be harnessed to make smarter trad- ing decisions by extracting information from hundreds of sources; for pipelinemonitoring, for identifying the likes of encroachments, rusting, and leakage and to conduct maintenance and inspection via drones, for example; back office functioncompliance, using robot-basedautoma- tion to reduce overall costs. Building the modelling context Infusing AI into one’s operations is but one aspect of a coordinated digital transformation strategy, but for those forward-looking players that take advantage of AI, cognitive computing and deep learning, the benefitswill be improved asset health and more efficient operations. Building the data models that enable AI takes considerable time – masses of historical data records need to be gathered to create the ‘context’ in which today’s data can be analysed and operations optimised. For this reason, it’s important to start one’s AI journey sooner rather than later. In the future, the ability to blend the power of artificial intelligence with human intuition and creativity will be essential in navigating an uncertain future that’s characterised by fewer sourcesof growth, slimmermargins, and increas- ing complexity. q

September 2017 • MechChem Africa ¦ 35