For GE, Machine Learning will enable them to optimize their design of wind farms. This serves two purposes, one is the financial benefits captured from improved operations and the other is environmental benefits of clean energy. The financial benefit for GE is a 20% boost in energy production resulting in $100 million is savings over the lifetime of a 100MW farm.3<\/sup> The environmental benefits are also significant, as we continue to add efficient clean technology assets that are cost competitive and attractive from an investment standpoint. This will create a virtuous cycle effect, driving further investment in technology and application that will serve to further reduce the cost of renewable energy. GE can have a huge impact on this industry, as they are currently 3rd<\/sup> in global market share with 60.4GW of installed capacity in 2016 4<\/sup> (Exhibit 1). Virtual simulators are necessary for process optimization and industry education, a requirement for reaching the U.S. Department of Energy’s \u201c20% Wind Energy by 2030\u201d 2<\/sup>.<\/p>\n To further reduce costs through operational efficiency and enable GE to continue competing in the power industry, continued investment in their Predix software and machine learning models is critical. Their short-term objective is to continue building out the cloud models and designing efficient wind farms before ever constructing the first physical wind turbine. In the medium term, GE\u2019s vision revolves around the industrial internet of things. The industrial internet of things involves the widespread distribution of sensors throughout industrial processes.3<\/sup> These sensors will constantly feed data to each other and share their information to make decisions on process parameters. Machine learning, incorporated with human oversight and training, will enable these industrial systems to optimize performance through automation with minimal costs.<\/p>\n While this is a great application of machine learning, there are many other ways GE can continue to use new technologies to drive operational efficiencies. One short term instance of this would be to take the same data used to optimize efficiency and apply it to turbine design. As you increase the number and types of parameters the machine learning program can play with, you can also enable it to suggest updates to design that will result in cheaper or more efficient turbines. In the medium term, the industrial internet of things will likely spread out of the generation plant and into the grid. This will enable GE to clearly understand instant pricing and demand of electricity. Machine Learning can be used to forecast future demand based on patterns in human behavior. This demand forecasting will help GE to determine the optimal time to release stored energy onto the grid or potentially stall operations for maintenance activities.<\/p>\n As a final thought, I\u2019d like to pose a question for reflection. The world\u2019s electricity demand will grow by 50% over the next 20 years. What are some other ways we can utilize technology in regulatory, operational, political and scientific applications to ensure that we are providing reliable, affordable and sustainable power?<\/p>\n (Word Count 754)<\/p>\n Exhibit:<\/p>\n Exhibit 1: Global Wind Generation Market Share 4<\/em>.<\/p>\n Endnotes:<\/p>\n 1 \u00a0<\/sup>CIOReview, \u201cGE’s Digital Wind Farm: The New Connected and Adaptable Wind Energy Ecosystem,\u201d CIOReview.com, May 26, 2015, https:\/\/www.cioreview.com\/news\/ge-s-digital-wind-farm-the-new-connected-and-adaptable-wind-energy-ecosystem-nid-6387-cid-53.html, accessed November 2018.<\/p>\n 2 <\/sup>Do, P. T. et al. (2013) \u2018Effects of 3D Virtual Simulators in the Introductory Wind Energy Course: A Tool for Teaching Engineering Concepts,\u2019, Comprehensive Psychology. doi: 10.2466\/04.07.IT.2.7.<\/p>\n 3<\/sup> Tomas Kellner, \u201cWind in the Cloud? How the Digital Wind Farm Will Make Wind Power 20 Percent More Efficient,\u201d https:\/\/www.ge.com\/reports\/post\/119300678660\/wind-in-the-cloud-how-the-digital-wind-farm-will-2\/, accessed November 2018<\/p>\n 4<\/sup> WindPowerMonthly, \u201cTop ten turbine makers of 2017,\u201d Windpowermonthly.com, October 2, 2017, https:\/\/www.windpowermonthly.com\/article\/1445638\/top-ten-turbine-makers-2017, accessed November 2018.<\/p>\n","protected":false},"excerpt":{"rendered":" Learn about how GE is using machine learning and virtual power plants to develop the next generation of optimized wind farms, driving down the cost of renewable energy to meet sustainability goals!<\/p>\n","protected":false},"author":11284,"featured_media":0,"comment_status":"open","ping_status":"closed","template":"","categories":[26,346,2373,1507],"class_list":["post-29032","hck-submission","type-hck-submission","status-publish","hentry","category-clean-energy","category-machine-learning","category-process-improvement","category-windfarm","hck-taxonomy-organization-ge","hck-taxonomy-industry-energy","hck-taxonomy-country-united-states"],"connected_submission_link":"https:\/\/d3.harvard.edu\/platform-rctom\/assignment\/rc-tom-challenge-2018\/","yoast_head":"\n![](\"http:\/\/offlinehbpl.hbpl.co.uk\/NewsAttachments\/MPW\/OEMchart.jpg\")
<\/p>\n