Article by Steve Long, General Manager Asia Pacific and Japan, Intel
Over the past few decades, the world has witnessed the rise of Asia Pacific through its manufacturing capability, fueled by labor, economic and regulatory advantages. However, the world’s manufacturing powerhouse is now at a critical inflection point.
Ever-changing customer expectations and demand for faster time-to-market, coupled by soaring inflation, supply chain disruptions and production backlogs are putting a strain on the region’s production. Not only are manufacturers facing challenges to keep up, but there is mounting pressure to evolve as new manufacturers spring up with the latest equipment and technologies.
To increase competitive advantage, many manufacturers in Asia Pacific are hoping to capitalise on “smart manufacturing” – the concept of integrating technologies, data, processes, and human interactions to improve production results.
Yet, there are two key barriers.
First, many manufacturers assume they have achieved smart manufacturing by implementing technologies such as artificial intelligence (AI) or data analytics in a piecemeal manner, limiting the benefits to the production floor instead of connecting them to the wider business value chain. Second, many are also hesitant to adopt new technologies due to concerns of interoperability between systems, large capital investments and the inability to scale.
The key challenge that contributes to these barriers is the management and integration of “data” and “processes”. It is not just about adding new technologies, organisations must realise true convergence of their factory (or operational technology) and enterprise (or information technology), run factories in a software-defined manner, and look at operations, workflows and human interactions in a holistic manner with the wider business context.
Realising true convergence of operational technology (OT) and information technology (IT)
Traditionally, OT and IT have operated separately.
OT like the machines and equipment in the factories are neither networked nor interconnected. They are often proprietary, verticalised systems that operate in silos. In most cases, human operators are required to monitor and manage the programming and physical operations of each equipment due to a lack of common standards across the machines. Think of a car factory, where the general assembly line has no visibility of the welding station that comes before it. They do not have the common “language” to “talk” to each other.
Now, thanks to the advances of internet of things (IoT), machine-to-machine communication and data analytics, the two worlds of OT and IT are finally converging. IT from the enterprise perspective is breaking the OT information silos by sharing and processing data exchanges across the manufacturing floors, enabling better productivity, automation and streamlined workflows.
Although it is a good start that some manufacturers are implementing new technologies to improve their factories’ productivity, many of them stop short at this stage without unleashing the full benefits of OT-IT convergence that can reach beyond the manufacturing floors. They overlook the ultimate impact in profit and loss (P&L) from a business perspective, be it from the view of business use cases such as return on assets (ROA), quality or yield optimisation, and production optimisation.
To achieve true OT-IT convergence, further integration of manufacturing data from these technologies with business intelligence such as enterprise resource planning (ERP) or supply chain management systems are needed, which further influence the decision-making outside of the factory, be it in supplier management, accounting or compliance.
New technologies and interoperability of machineries are only the first mile of the smart manufacturing journey. The steps that follow are the differentiator that will connect factories to the business as a cohesive whole and truly create competitive advantage.
Software-defined manufacturing – when manufacturing plants operate like an IT system
Traditional manufacturing has long been defined by hardware, where individual pieces of equipment are designed for one repetitive task. When it comes to making changes, it can require huge amounts of capital to upgrade the factory. This is where software-defined manufacturing comes in.
Software-defined manufacturing means software can configure, monitor, and manage machines and their processes across the manufacturing floor. It allows manufacturers to do more with the existing hardware and enable one hardware to have multiple functions or pivot for other purposes. Think of how one smartphone has replaced the phone, camera, and GPS in one single device. The same is now happening in the manufacturing world, where manufacturers can operate their factories like an IT system.
This enables increased flexibility and faster programming to both individual machines and the entire production process through just one interface. Manufacturers can also virtualise physical machines to create digital twins in an on-premise or cloud environment to simulate how an upgrade would impact a production line. With AI and machine learning across the manufacturing edge, data can be analysed closer to where it is collected, and adjustments can be made in near real-time to optimise operations.
Yet, one thing that many manufacturers overlook is that software-defined manufacturing enables you to constantly update and upgrade. Therefore, manufacturers should not stop exploring new technologies, workflows and processes once an upgrade is made. It is through continuous experimentation and adjustments to the facilities can manufacturers benefit their businesses in the long run.
Building the architecture for the future of smart manufacturing
To make a smart manufacturing future possible, it is important to have an underlying architecture that simplifies OT-IT convergence and enables software-defined manufacturing. Manufacturers need a foundation that makes it possible to design, scale and run separate functions on one united, cloud-like platform.
This foundation requires hardware and software building blocks to consolidate disparate functions including process control, visualisation and data acquisitions. This also means having the right silicon, enhanced for demanding industrial applications that can converge different applications that would otherwise require multiple CPUs, GPUs and accelerators.
That said, the future of smart manufacturing can only be successful if the entire manufacturing ecosystem, including the original equipment manufacturers (OEM), factories, system integrators and more are able to integrate technologies, data, processes, and human interactions together. This requires every area of the manufacturing ecosystem to embrace an open, fully programmable, and standards-based unified system, so manufacturers have the choice, flexibility and interoperability to optimise operations and drive innovation regardless of the vendors or suppliers they use.
This, is the future of smart manufacturing.
