Ir para o conteúdo
Blogs de indústria da Microsoft

So promising is the Internet of Things (IoT) for the Oil & Gas industry, the key question is not “if” or “when”, but how to get the ball rolling, and, of course, rolling in the right direction. After discussing the specific challenges in The “Internet” AND “Things” for Oil & Gas, we now address how to move from the concept of this disruptive technology to implementation? A two-part article by Gartner, “A CIO’s Guide to Realizing the Business Value of IoT, 08 July 2015, Chet Geschickter” provides five steps that, in my experience, almost guarantee success.

02_azure_iot_oil_gas_mid

 

Step 1 — Determine the Strategic Rationale:

Especially in today’s Oil & Gas world where every IT investment dollar counts, finding the right use cases and quantifying the value in hard cash returned to the business is a must. My hardline recommendation would be: Don’t even kick off the project unless you know the bottom line.

As an example: drilling involves many systems and inherent risks such as stick-slip which can damage the rig, causing nonproductive time and potential safety incidents. The NuGreen Drilling solution, developed with OSIsoft and Element Analytics, uses advanced machine learning models to predict stick-slip and all other risks, while optimizing rate of penetration, which means the well will be drilled as quickly as possible. The cost saving for an onshore rig would be multiple hundreds of thousands of USD per day, easily justifying the cost of this predictive analytics investment.

Read more in Joseph Sirosh’s keynote at the Global Energy Forum discussing the topic “From the Well to the Cloud: Advanced Analytics for the 21st Century”.

Step 2 — Create IoT Use Cases:

We’ve seen great results in brainstorming sessions with line-of-business and IT leaders, where we collect even the wildest ideas and prioritize the top 2-3 use cases based on criteria such as complexity of solutions, business problems addressed, cost/risk of doing nothing, availability of implementation resources and sponsors, etc. This joint IT and Business exercise drives both critical aspects of Industrial IoT, the business value and feasibility of success. Furthermore, such exercises are often the initial spark for an innovation breakthrough, smashing through past paradigms and preconceived obstacles within an organization.

Step 3 — Set Solution Scope to Make Use Cases Feasible:

As engineers we often try to boil the ocean; our DNA is predisposed to build the perfect solution. However, in the accelerated digital age, we must be more focused on delivering a manageable set of capabilities which deliver a significant portion of overall value. Baker Hughes FieldPulse is a good example of this approach. The low oil price is redirecting operators to invest further in technology to reduce lifting costs, increasing recovery and operational efficiency. Operators need standardized, cost-effective and repeatable Digital Oil Field (DOF) technology. The technology needs to be quick to deploy at a low total cost of ownership and capable of being deployed across a broad portfolio of assets. Historical solutions have typically been very customized, expensive and take too long to implement.

In contrast, FieldPulse, automates six fundamental workflows that drive well performance from field data capture to production forecasting, to well and field KPIs using dynamic boundaries and model-based tolerances. The implementation of this solution can start with a very low cost per well, leveraging pure OPEX budget, without significant infrastructure investments. The first well can be implemented in hours, can be repeated on any well of any type, and is easily scalable from 1 well to 1000 wells.

03_azure_iot_predictive_maintenance

 

Step 4 — Establish Deployment Options and Technology Sourcing:

The Microsoft Azure IoT Suite successfully follows a phased implementation methodology to strictly manage risks. The starting point is often data ingestion, which can be one of the biggest challenges due to a multitude of data silos, legacy applications with closed APIs and “things” that use many different protocols to expose data. Luckily, in many cases there is already an infrastructure in place to access, collect and store a significant portion of field data, giving the Oil & Gas industry a head start.

Getting the IoT data and storing it in the cloud does have its unique challenges. The data transfer pattern is often lots of small pieces of data with high velocity. In some cases, high volume batch is needed. Microsoft recently introduced the Data Lake, which is an infinitely large and scalable platform able to manage PB size files. Data Lake is built on Microsoft’s internal big data store, which runs all of the web assets and currently holds Exabytes of data. For real time analytics, Azure Stream Analytics services as well as open source projects such as Kafka can be used. Batch analytics can be executed with Azure Machine Learning, HDInsight and the new Data Lake Analytics service, as well as Apache Spark.

Lastly, the business transformation that is driven by IoT impacts an organization’s business processes, which sometimes can’t be delivered by legacy applications. Fortunately, there are new generation business platforms, such as Microsoft Dynamics, to support these disruptive business processes, which should be considered in the early planning phases of IoT projects. “The reason why it is so difficult for existing firms to capitalize on disruptive innovations is that their processes and their business model that make them good at the existing business actually make them bad at competing for the disruption.” — Clayton Christensen, Author of The Innovator’s Solution and The Innovator’s Dilemma.

Step 5 — Conduct Pilot Projects to Validate Technology and Verify Benefits Capture:

We know that seeing is believing. So, in our conversations with customers, one of the first things we do is recommend a pilot. The best part of this is that with the Azure IoT platform you have the freedom to test a lot of ideas without significant investments, efforts or risks. Furthermore, you are always able to switch between use cases, models and architectures without major financial nor time impact.

As we’ve previously mentioned, probably the only part of an IoT project which may need up-front investment are incremental physical devices and architectures for data ingestion. However, in our standard approach, integrating with your existing Operating Technology (OT) infrastructure almost eliminates this cost. The best way to start is by leveraging what you already use for other purposes, connect it to Azure IoT for value-add services, and start testing new capabilities like machine learning for predictive maintenance. For instance, if you already have an OSIsoft PI infrastructure, it is just a matter of setting the right security principles and architecture, while connectivity to the Azure IoT platform is achieved with the PI Cloud connector. So, in only a few clicks, you have access to the full Azure IoT Suite capabilities.

A good example of the successful phased approach is the work done by our partner Avanade, for Freeport-McMoRan, one of the largest international resources companies. The initial IoT application to monitor the entire mine site was completed in a few months, quickly yielding business value by reducing equipment downtime and overall production costs. Mine supervisors were delivered a single pane of glass to monitor and react in near-real-time to important KPIs, thereby optimizing production processes across the mine sites, instead of making sub-optimal decisions due to fragmented data, disconnected applications and processes.

The bottom line that all of our early adopters have experienced, is that the combined agility and capability of advanced technologies within the Azure IoT Suite, as well as ease of deployment can quickly deliver measurable business value.