AI: The connective tissue in the digital business
By supercharging the benefits of other technologies—including augmented reality (AR) and digital twins—AI can transform operational performance.
Connectivity for frontline workers, whether it’s between automated manufacturing processes or between healthcare specialists, has already made businesses more efficient. New research from TeamViewer finds that 84% of businesses agree that technology connectivity offers them a real-time overview of operational performance and challenges. A similar number say it increases collaboration across functions.
Now, AI is taking connectivity to the next level.
AI applications can process vast amounts of data from multiple sources at speed. For businesses looking to reduce downtime, automate simple tasks, and accelerate problem-solving, the operational benefits are enormous.
The implications for IT infrastructure are deeper still. Gartner predicts that, by 2027, GenAI tools will be used to explain legacy business applications and create appropriate replacements, reducing modernization costs by 70%.
AI supports upskilling in real time
But, as with so many technologies, the true power of AI only really becomes apparent when it is infused with other technologies.
In the words of Mei Dent, Chief Product and Technology Officer at TeamViewer, AI is the invisible engine. When AR is used on an automotive assembly line or to repair complex machinery, for example, AI can act as a real-time assistant.
“AI can reduce the cognitive load on employees and allow them to work – and augment their knowledge – in the moment, with information at their fingertips,” says Stefan Baumgart, Director Product Management at TeamViewer. Such access could be transformative, especially when nearly two-thirds of frontline workers say they spend too much time searching for information.
“I can have AI read every manual out there and ask, for example, ‘What torque value do I use on the crank casing?’,” adds Baumgart. “You are bringing in a structure that understands the context and the domain.”
As AI starts to decipher streams of data and analyzes a constantly evolving set of digital signals, it will have applications for more than problem-solving, says Baumgart. Workers can embed it in their workflows, whether that is a field support manual designed to be consumed by AI, or AI-supported design of changeover processes on manufacturing lines.
Predictive monitoring reduces downtime
A shift to a proactive, data-driven approach to maintenance based on 24/7 monitoring of equipment, plus data acquisition and trend analysis, reduces equipment failure and consequent downtime.
It also gives companies visibility of other likely issues, such as bottlenecks, before they happen.
It can also help to solve recurring issues. For more than 10 years, Shell’s Perdido floating oil and gas platform experienced periodic disruption of the pumps responsible for separating oil and gas.
When Shell used AI to analyze the pumps’ operational performance, it discovered a chemical signature in about 70% of the cases of pump disruption. Proactively looking for this signature allows the company to predict and mitigate future disruption.
Increased connectivity is also invaluable in this kind of remote industrial operation. If AI can continuously scan machinery and equipment for signs of faults and errors and alert human engineers, there is greater potential to resolve the issue remotely.
Remote connectivity and digital twins provide essential intel
Product design is another area that can be accelerated with AI. Companies can use digital-twin simulations to predict how a product or process will perform. This information can then influence physical design. AI can help order and interpret the vast volume of data that digital twins need. McKinsey also suggests that GenAI can “supplement data training sets used by digital twins by creating synthetic data.”
Sacha Porges, Global Director for Customer Quality and Programs at GKN Automotive, offers another example. GKN is investigating the use of digital twins for design and development work—building virtual versions of new projects that can then be tested, rather than having to depend on physical prototypes.
“One of our biggest customers effectively wants 90% of our design validation testing to be conducted virtually. Rather than building a very expensive set of prototypes, we can simulate the prototypes to perform the majority of the necessary testing to minimize costs and time.”
Global Director for Customer Quality and Programs at GKN Automotive
Complexity and security are the main worries
So, what is stopping organizations from embedding AI more widely?
In a Deloitte survey, 93% of businesses said they believed that AI will be a pivotal technology to drive growth and innovation in the manufacturing sector. But only half of the companies in McKinsey research said they used AI in two or more functions.
Complexity and competitive advantage are two considerations. “As a global company, we require standardization,” says GKN’s Sacha Porges. “There are multiple vendors with many interesting solutions, but the ideal solutions will be scalable quickly and simply so they can be deployed globally throughout our footprint. We also need partners that are willing to support us to learn these systems for ourselves so we can achieve an acceptable level of autonomy.”
Another concern for businesses is data security and how AI is used to analyze large data sets. Collaborating with third parties and suppliers can take proprietary data outside of the organization’s usual boundaries and open up a bigger attack surface, making the business more vulnerable to malicious data manipulation – known as data poisoning.
But despite these barriers, there is excitement about an “invisible” technology having very visible benefits on the frontline of company operations.