BUSINESS PROBLEM-SOLVING CASE GE Bets on the Internet of Things and Big Data Analytics General Electric (GE) wants to be known as the worid's Digital Industrial Company. It is moving away from traditional manufacturing toward a much more technology-centric business strategy and busi- ness model that focuses on electric power genera- tors, jet engines, locomotives, and oil-refining gear and software to connect these devices to the cloud. GE is putting its money on the technology that con- trols and monitors industrial machines as well as software-powered, cloud-based services for analyz- ing and deriving value from the data. GE hopes this strategy will turn it into a major software company. GE is using sensor-generated data from industrial machines to help customers monitor equipment performance, prevent breakdowns, and assess the machines' overall health GE has committed $l bil- lion to installing sensors on gas turbines, jet engines, and other machines' connecting them to the cloud; and analyzing the resulting data to identify ways to improve machine productivity and reliability. ]n other words, GE is betting its future on software and the Internet ofThings (IoT). In a number of industries, improving the produc- tivity ofexisting assets by even a single percentage point can generate signicant benefits. This is true of the oil and gas sector, where average recovery rate ofan oil well is 35 percent. That means 65 percent of a \"Ell-S potential is left in the earth because avail- able technology makes it too expensive to extract If technology can help oil extraction companies raise the recovery rate from 35 to 36 percent, the world's output would increase by 30 billion barrelsthe equivalent of three years of global supply. The oil and gas industry is also deeply affected by unplamred downtime, when equipment cannot oper- ate because ofa malfunction. A single unproductive day on a platform can cost a liquefied natural gas {LNG} facility as much as $25 million, and an aver- age midsized LNG facility experiences about five dolwn days a year. That's 3125 to $150 million lost. Minimizing downtime is critical, especially consider- ing declining revenues from lower energy prices. GE sees a $1 billion opportunity for its IoT software. The foundation for all ofGE's Industrial Internet {IoT} applications is Predix, a software platform launched in 2015 to collect data from industrial sen- sors and analyze the information in the cloud. Predix can run on any cloud infrastructure. The platfonn has open standards and protocols that allow customers to more easily and quickly connect their machines to the Industrial Internet. The platform can accom- modate the size and scale ofindustrial data for every customer at current levels of use, but it also has been designed to scale up as demand grows. Predix can offer apps developed by other companies as well as GE, is available for on-premises or cloud-based deployment and can be extended by customers with their ow'n data sources, algorithms, and software code. Customers may develop their own custom applications for the Predix platform. GE is also building a developer community to create apps that can be hosted on Predix. Predix is not limited to industrial applications. It could be used for analyz- ing data in healthcare systems, for example. GE now has a Health Cloud running on Predirt Data security is embedded at all platform application lay- ers, and this is essential for companies linlring their operations to the Internet. GE currently uses Predix to monitor and maintain its own industrial products, such as wind turbines, jet engines, and hydroelectric turbine systems. Predix is able to provide GE corporate customers' machine operators and maintenance engineers with real-time information to schedule maintenance checks, improve machine efficiency, and reduce downtime. Helping customers collect and use this operational data proactively would lower costs in GE service agreements. When GE agrees to provide service for a customer's machine, it often comes with a performance guarantee. Proactive identification of potential issues that also takes the cost out of shop visits helps the customer and helps GE. In early 2013, GE began to use Predix to analyze data across its fleet ofmachines. A single engine's operating data will only tell you there's a problem with that engine. But by collecting massive amounts of data and analyzing the data across its entire fleet ofmachines, GE was able to cluster engine data by operating environment. The company found that the hot and harsh environments in the Middle East and China caused engines to clog, heat up, and lose efficiency, so they required more maintenance. GE found that engines had far fewer ofthese problems if they were washed more frequently. Fleet analytics helped GE increase engine lifetime and reduce engine maintenance. The company thinks it can save its customers an average of $7 million ofjet airplane fuel annually because their engines will be more 423' efficient. Predix's robust data and analytics platform risk-assessment solution combines internal and exter- made it possible for GE to use data across every GE nal factors (such as flooding) to provide an accurate, engine all over the world and cluster fleet data. up-to-the minute visual representation of where risk Predix is starting to provide solutions for GE cus- exists in a pipeline. This risk assessment tool enables tomers. For example, Invenergy LLC, North America's pipeline operators to see how recent events affect their largest independent, privately held renewable energy risk and make real-time decisions about where field provider is implementing GE Asset Performance Man- service crews should be deployed along the pipeline. agement reliability management software on 13 tur- The risk assessment tool visualization and analytics bines at six gas-operated thermal plants in the United capabilities run on Predix. States. The software is based upon Predix and will Weather has a sizable impact on risk for pipelines in perform predictive analytics on turbines that produce areas prone to seismic activity, waterways, and wash- an operating capacity of 3,159 megawatts, enough to outs. Checking weather patterns along thousands of power a half-million homes. The software will help miles of pipe for rain or flood zones, and integrating plant technicians more accurately predict and diagnose those data with other complex pipeline data sets is equipment failures before they occur in order to avoid very difficult to perform manually. But by bringing all unplanned outages. During preliminary testing, the relevant data together in one place, GE gives pipeline system identified a turbine journal bearing experience operators easier access to information to help them ing early-stage vibration. A failure could have damaged address areas with the greatest potential impact. the turbine and produced an unplanned outage, but GE expects customers to benefit immediately from Invenergy technicians were able to detect and repair having all of their data integrated. But it wants them to the malfunction three months before it would have be able to do more. In addition to being able to examine occurred. The GE system will also help Invenergy bet- all current risk, pipeline operators would benefit from ter manage its assets to achieve the optimal solution for a "what-if" calculation tool to model hypothetical sce- reducing cost, managing risk and improving availabil narios, such as assessing the impact of adjusting operat- ity and reliability. ing pressures or addressing particular areas of corrosive British oil and gas company BP pic had been using its pipe. GE would give them the tools for a color-coded own software to monitor conditions in its oil wells but view of how those actions affect pipeline risk. decided to get out of the software business and became GE wants to go beyond helping customers manage a GE customer. By the end of 2015, BP equipped 650 the performance of their GE machines to managing the of its thousands of oil wells with GE sensors linked to data on all of the machines in their entire operations. Predix. Each well was outfitted with 20 to 30 sensors Many customers use GE equipment alongside of equip- to measure pressure and temperature, transmitting ment from competitors. The customer cares about run- 500,000 data points to the Predix cloud every 15 sec- ning the whole plant, not just GE turbines, for example, onds. BP hopes to use the data to predict well flows and 80 percent of the equipment in these facilities is not and the useful life of each well and ultimately to obtain from GE. If, for example, if an oil and gas customer has an enterprise-wide view of its oil fields' performance. a problem with a turbo compressor, a heat exchanger GE identified pipeline risk management as a major upstream from that compressor may be the source of challenge for the oil and gas industry. There are 2 mil- the problem, so analyzing data from the turbo com- lion miles of transmission pipe throughout the globe, pressor will only tell part of the story. Customers there- moving liquid oil or gas from its point of extraction to fore want GE to analyze non-GE equipment and help refining, processing, or market. About 55 percent of them keep their entire plant running. transmission pipeline in the United States was installed If a customer purchases a piece of GE equipment before 1970. Pipeline spills are not frequent, but when such as a gas turbine or aircraft engine, GE often they occur, they cause serious economic and environ- enters into a 10- to 15-year contractual services agree- mental damage as well as bad publicity for pipeline ment that allows GE to connect to and monitor that operators and energy companies. Pipeline operators machine, perform basic maintenance and diagnostics, are always anxious to know where their next rupture and provide scheduled repairs. GE receives a bonus will be, but they typically lacked the data to measure payment for keeping the equipment running at a pipeline fitness. Operators had no way of integrating specified threshold. GE may now be able to apply multiple sources of data into one place so they could such outcome-based pricing to coverage of non-GE see and understand the risk in their pipelines. machines. GE developed a pipeline-management software suite GE CEO Jeffrey Immelt wants GE to become a top for accessing, managing, and integrating critical data 10 software company by 2020. In order to do this, GE for the safe management of pipelines, including a risk needs to sell vast amounts of applications and Predix- assessment tool to monitor aging infrastructure. GE's based analytics. Although few businesses have therefining, processing, or market. About 35 percent of them keep their entire plant running. transmission pipeline in the United States was installed If a customer purchases a piece of GE equipment before 1970. Pipeline spills are not frequent, but when such as a gas turbine or aircraft engine, GE often they occur, they cause serious economic and environ- enters into a 10- to 15-year contractual services agree- mental damage as well as bad publicity for pipeline ment that allows GE to connect to and monitor that operators and energy companies. Pipeline operators machine, perform basic maintenance and diagnostics, are always anxious to know where their next rupture and provide scheduled repairs. GE receives a bonus will be, but they typically lacked the data to measure payment for keeping the equipment running at a pipeline fitness. Operators had no way of integrating specified threshold. GE may now be able to apply multiple sources of data into one place so they could such outcome-based pricing to coverage of non-GE see and understand the risk in their pipelines. machines. GE developed a pipeline-management software suite GE CEO Jeffrey Immelt wants GE to become a top for accessing, managing, and integrating critical data 10 software company by 2020. In order to do this, GE for the safe management of pipelines, including a risk needs to sell vast amounts of applications and Predix- assessment tool to monitor aging infrastructure. GE's based analytics. Although few businesses have the Chapter II: Improving Decision Making and Managing Knowledge 429 capital or infrastructure to operate a platform for inte- Sources: Paul Gillin, "Industrial loT Revs Up at Big Renewable Energy grating and analyzing their loT data, GE faces compe- Provider," Silicon Angle, May 17, 2017; www.predix.do, accessed May 20 2017; Laura Winig, "GE's Big Bet on Data and Analytics," MIT Sloan tition from many sources. Amazon, Google, IBM, and Management Review, February 2016; www.ge.com, accessed May 20, 2017. Microsoft are all getting into Internet of Things plat- Devin Leonard and Rick Clough, "How GE Exorcised the Ghost of Jack Welch to Become a 124 Year-Old Startup," Bloomberg Businessweek, March forms, and dozens of start-ups have similar ambitions. 21, 2016; Holly Lugassy, "GE Leverages Pivotal Cloud Foundry to Build The biggest question is whether other large industrial Predix, First Cloud for Industry," Cloud Foundry.org, May 11, 2016; Cliff companies will turn to GE or to another cloud plat- Saran, "GE Predictive Analytics Optimises Irish Power Electricity Pro- duction," Computer Weekly, July 13, 2015; Charles Babcock, "GE Predix form to manage their information. And if you're a Cloud: Industrial Support for Machine Data," Information Week, August 6, manufacturer of some size and sophistication, will you 2015; and "GE: IoT Makes Power Plants $50M More Valuable," Information allow GE to "own" the data on your business, or will Week, September 29, 2015 you manage and analyze the data yourself? CASE STUDY QUESTIONS II-13 How is GE changing its business strategy and value to the firm of each of those decisions? business model? What is the role of informa- Explain. tion technology in GE's business? 1 1-16 To what extent is GE becoming a software 11-14 On what business functions and level of deci- company? Explain your answer. sion making is GE focusing? 1I-17 Do you think GE will become one of the top 1 1-15 Describe three kinds of decisions that 10 U.S. software companies? Why or why not? can be supported using Predix. What is the