Imagine if doctors could create a digital twin version of you that updates in real time using data from sensors in your home, your car and on your person – such as wearable devices like smart watches. This virtual twin could signal or even predict impending illnesses such as cancer, enabling early diagnosis when treatment might be most effective.
Digital twin data technology is already deployed to monitor the operational health and functionality of some of the world’s most complex and mission-critical machines. Digital twin versions of jet engines, locomotives and gas turbines are tracking wear and tear on the machinery, maximizing operational efficiency and predicting when these machines will need maintenance – often before they break down.
Using exact duplicates to manage complex systems dates back to NASA’s moon missions of the 1960s and 1970s. NASA used mirrored systems, the precursor of digital twins, to rescue the Apollo 13 mission when it ran into trouble. Nearly 50 years later, digital twin technology helps NASA understand and manage the operation of machines that are moving through the vast reaches of our solar system.
Digital twin technology uses cloud-connected sensors embedded in machines to upload real-time operational data, producing up-to-date virtual simulations of real-world machines. Manufacturers can then use edge analytics to analyze and evaluate how their products are performing in the field. The ultimate goal: Have a digital twin running for every real-world asset in the field, with the digital twin updating its status as it receives operational data.
The Internet of Things (IoT) is key to the implementation of digital twin technology. The increasing affordability of sensors, widespread use of WiFi and the data-throughput capacity of the cloud combine to make the application of large-scale digital twin modeling affordable for a range of manufacturers operating in the industrial IoT (IIoT).
When manufacturers can see real-time data on how their products are operating, they can make dramatic improvements in design, innovation, efficiency and manufacture. That capability enables them to proactively contact end users so plans can be made for repairs or maintenance – heading off the disruption of potentially costly breakdowns.
5 Steps for Turning Industrial IoT Data into a Competitive Advantage
Read about the five steps SAS and Intel say will help you craft a strategy for turning raw IIoT data into valuable insights – from defining business goals and analytics strategy to choosing the right platform.
Analytics: Transforming IoT data into value
Bryan Saunders, principal industry consultant for the global IoT practice at SAS, points out that SAS has deep experience in the analytics space. “We’ve been in that game for 40-plus years.
“The heart of the digital twin is the analytics,” he says. “It’s not just about, ‘Can you collect the data,’ but can you turn it from data to valuable transformative information? The main driver for that is analytics.
“This means you have to be able to collect and move the data in effective ways. Then you have to understand what it’s telling you. But beyond that, you have to drive the action so that you can achieve that expected result on the back end.”
Saunders concentrates on the industrial side. “My work is focused on using heavy industrial-connected assets to drive improved availability, efficiency, safety and reliability in the energy and manufacturing sectors. For example, take a heavy-duty gas turbine. Being able to effectively understand the baseline position of that asset, I can look at how it has failed historically and ways that I think it can fail in the future and use those analytical techniques to provide predictions for maintenance activity so that you don’t have a catastrophic failure.
“When you have visibility into what the future may hold, it drives significant efficiencies. Real-time anomaly detection and health assessments have proven extremely valuable in terms of predictive maintenance. Studies show that by using the data in that fashion, you’re reducing unplanned maintenance by up to 80 percent.”
GE, a leader in virtual twin technology, is using a combination of artificial intelligence-driven analytics and visual sensors on matchbox-size robots to look for cracks inside working engines. Recent advances allow cameras to find cracks even on dirty or rusty turbine blades. Visual sensors on drones can inspect for corrosion on the 200-foot tall stacks that burn off excess gas at oil and gas production sites.
Tesla Motors is another example of a company that is deeply invested in digital twin technology to provide better service and reliability for car owners. Tesla creates a digital twin of every car it sells. Tesla then updates software based on individual vehicles’ sensor data, and uploads updates to its products. This data-driven software development process enables more efficient resource allocation and a markedly better user experience for the vehicle owner.
At its factories, Bosch is comparing sensor-driven production data to a digital twin of production lines running at 100 percent efficiency. As a result, production deviations can be quickly flagged and trends can be more easily identified. These smart, connected production lines have enabled a 25 percent output improvement in the company’s electronic stability program and automatic braking systems.
Saunders sees IoT and analytics as being transformative in four major sectors: Smart cities, connected vehicles, connected factories and smart grid or utilities applications. All these transformations ultimately create value for connected consumers. For Saunders, IoT and analytics have proven to be an indispensable competitive edge in all phases of a product’s life cycle, from design to manufacturing, maintenance and operation.
As he explains: “A wide range of industries are coming to understand this is the way we have to do business to be successful and to stay competitive in this new world of IoT.”
Daniel Newman talks IoT: Smart factories
Listen as Daniel Newman and Marcia Walker of SAS discuss how the IoT is affecting the future of the smart factory. Using IoT data, for example, manufacturers can achieve dramatic improvements in manufacturing quality – for processes, equipment and productivity.
- ModelOps: How to operationalize the model life cycleModelOps is where analytical models are cycled from the data science team to the IT production team in a regular cadence of deployment and updates. In the race to realizing value from AI models, it’s a winning ingredient that only a few companies are using.
- Anti-money laundering and counter-terrorist financingTo monitor for financial crimes, an essential element for both anti-money laundering and counter-terrorist financing is using data analysis to detect unusual activity during processing by monitoring transactions, customers and the network of behaviors.
- Contextual engagementCreating contextual customer engagements means understanding enough about each customer to ensure that all communications over time are progressively more anticipated, personal and relevant.