Digitisation is the new wave sweeping the industrial world. Cost of computing and communication has reached a point of inflection enabling application of this technology everywhere. Dramatic productivity improvements and disruption in business models have been observed in retail, banking, hospitality and transportation industries, for example, that have made these industries more competitive.
Industrial companies all over the world are grappling with how to integrate digital capabilities into their operations to get ahead of competition in terms of cost and driving growth through new services and new business models. Just how much of a difference could digital integration make? Some examples cited in literature include design cycles, which incur 20% faster product development cycle; manufacturing with 20% improvement in manufacturing and supply chain efficiency; operations with 20% improvement in output of a power plant and services with 10-20% reduction in maintenance cost and downtime of power plants.
To understand the kind of impact percentages like these can have on manufacturing operations, consider that each percentage point from a company represents upwards of $500 million in potential savings. For a large power plant, the improvement in operations and maintenance cost reduction can add up to $150 million/year.
Challenges in adopting digitalisation
The challenge experienced by today’s industries is identifying the right areas where digitisation will yield the desired results in productivity and growth. High volume automotive plants have significant automation and intelligence built into machines enabled by sensors that collect and transmit data for decision making. However, for low volume and older industries, which lack sensor-enabled automation, significant challenges exist in installing the right sensors to create digitisation-enabled productivity. Another challenge that industries are facing is identifying the right scale of application of cloud, Big Data, artificial intelligence and machine learning to leverage large volumes of generated data. A dearth of skilled engineers and software professionals, who can apply software and machine learning skills along with plant domain knowledge to provide reliable solutions to deliver bottom line impact, is another hindrance in the industry.
Digital Twin initiatives
Considering these challenges, GE is, thus, translating the power of the internet to products and services through Digital Thread and Digital Twin initiatives. GE is connecting design, manufacturing, operations and services of assets through Digital Thread technologies, so that key data required to improve productivity and operations is available throughout the life cycle of the asset.
Our company is using Digital Twin technology to model high-value assets through the asset life cycle to enable better and faster decisions and outcomes. Digital Twin is a dynamic digital representation of an industrial asset that allows industries to better understand, predict and manage the performance of their assets. This approach is enabling design productivity for assets such as jet engines, power turbines, wind turbines and locomotives. Through the Brilliant Factory initiative, we are optimising our factories to improve cycle time and reduce costs. Using Predix and asset performance management solutions, we are enabling reduced maintenance costs and optimised operations.
Advances in computing and communication technologies enable embedding of sensors into machines, leverage software to gather data and use artificial intelligence and machine learning to drive better insights and outcomes. This is called the Digital Thread, because the digital capabilities are horizontally weaved through our design, manufacturing and services facilities and vertically integrated through value chain, all enabled by Predix, the industrial cloud-based platform.
Brilliant Factory is an example of driving this vision of integrating design and manufacturing in our factories around the world. A select handful of our factories are already fitted with digital capabilities along with additive technologies and are realising productivity gains as a result. At a GE aviation facility in Auburn, Alabama, there are thirty 3D printers manufacturing fuel nozzles for jet engines—the first 3D-printed part ever approved by the FAA to fly in a commercial jet. At a GE Transportation facility in Grove City, Pennsylvania, a 10 to 20% reduction in unplanned downtime has been achieved through digitisation using Digital Thread. Additionally, in Florence, Italy, equipping machines with sensors has identified optimum times for maintenance with minimum disruption to production and even added an entire production line without adding a new shift.
Digital Thread enables closing the loop between design and manufacturing seamlessly. If during manufacturing an error is detected, the design intent is immediately reviewed and the next manufacturing process is adapted to give assurance that the part will meet the intent of design and expectations of the customer. This can be done in seconds, instead of taking days or weeks previously.
A Digital Twin is a dynamic digital representation of an industrial asset that enables industries to better understand and predict the performance of their machines. Industries are faced with preventing unscheduled failures to avoid significant losses in production while at the same time reducing cost of maintenance. Optimising the plant operations in the presence of changing environmental conditions and degrading equipment is another top focus for plant operators.
Digital Twin brings together machine learning and domain models used during design of industrial assets to provide insights that enable improved operations and reduced maintenance costs. Digital Twins provide a rich, constantly evolving picture of industrial assets capturing everything from components to functions to entire processes and plants. They capture the entire lifespan of an asset as well as entire asset classes and gain insights into the past and present performance and future intelligence. Furthermore, Digital Twins are an ideal software object to perform simulations, allowing for scenario testing and plant optimisation.
Digital Twins have been used in bearing anomaly detection for jet engines that provide 15 to 30-day heads-up on potential failures. This information can be used to prevent unscheduled downtime and optimise maintenance. Predictions for hot gas path parts in jet engines have enabled our customers to inspect and replace only those engines that require attention. This has resulted in $44 million in savings in engine inspection and maintenance costs. Using dynamic optimisation of maintenance schedules of jet engines using Digital Twins, which incorporate understanding of airline routes and maintenance shop availability, has resulted in $10 million annual savings by optimising the shop operations.
Through Digital Twin models, Evolution Locomotive minimises fuel consumption and emissions by considering the track conditions and locomotive tractive effort. Digital Twin models enable optimising full mission driving profile and this saves 32,000 gallons of fuel per locomotive and 174,000 tonnes in emissions per year. GE’s Renewables business is transforming the industry’s future with the launch of the Digital Wind Farm enabled by Digital Twins. The technology could boost a wind farm’s energy production by as much as 20% and create $100 million in extra value over the lifetime of a typical wind farm.
By better understanding performance of individual asset as well as fleets of assets, feedback can be provided to design processes so as to improve the design of these assets, extend the life of the components, improve maintenance practices and recommend upgrades and operational changes to customers, thereby, closing the loop of Digital Thread.
Huge gains in productivity
Digitisation of industrial plants and processes provides huge gains in productivity and delivers better outcomes to customers. Digital Thread and Digital Twin are two technologies that drive connectivity between design, manufacturing, operations and maintenance of industrial assets. Digital Twins are evolving digital models that combine domain/physics models with machine learning to provide better insights than was ever possible before. Multiple examples on improving productivity in design, manufacturing, operations and services have been provided to illustrate how we are using this technology to improve its operations and customer services. Digital transformation is a key opportunity to transform industries to be more competitive in the future.
The article is authored by Vinay Jammu, Technology Leader, Physical-Digital Analytics, GE Global Research and Vinod Kumar, Technology Leader, Materials and Manufacturing Technologies, GE Global Research