In 2018, the World Economic Forum (WEF) and McKinsey & Co. estimated that Industry 4.0 would offer manufacturers and suppliers $3.7 trillion in value-creation potential by 2025 . That was before industrial digital transformation underwent the COVID test of 2020. A McKinsey survey last year found that early adopters of Industry 4.0 technologies have fared better during the pandemic, although the bar for success in digitization has been raised from adding value to operating under trying circumstances. The pandemic has left crucial lessons in Industry 4.0 transformation.
Transformation driven by data
The predecessor of Industry 4.0 centered around the application of computers, the building of IT infrastructure, and using automation on the assembly line to transform manufacturing. This implementation of IT and semi-automated machines started to generate “dark data,” information that was ignored or erased until eventually tools became available to analyze it and reveal new intelligence.
The smart factories that Industry 4.0 enables is not merely about the hardware, such as advanced sensors, but about analyzing production data and combining it with similar intelligence all along the supply chain.
The previous industrial revolutions were built on steam, electricity, and then computers, but this one starts with data — gathering it where it is useful, sending it where it is needed, making sense of it all, and using new intelligence to effect better decisions. This is what enables exercising of control on manufacturing and other business processes to make them efficient and resilient.
Pillars of Industry 4.0
The four foundational pillars of Industry 4.0 are sensors, connectivity, data storage, and processing (Figure 1). These functional blocks may be part of various equipment, standalone devices, or embedded algorithms, forming a larger customized setup of a smart factory.
Sensors: Sensors are the point at which information is acquired and representative data is generated. They take measurements and help monitor processes. Sensor modules may be added to existing equipment or come embedded in newer smart machines.
Since they generate immense amounts of real-time data, and many issues at any one point in operations need immediate control, the use of smart sensors with on-board compute resources is increasing (Figure 2). For instance, applications like electronic component pick-and-place onto printed circuit boards benefit from measurement of variations and immediate corrections.
Connectivity: While sensors for process monitoring have long existed, Industry 4.0 expands the power of information by adding connectivity. The Industrial Internet of Things (IIoT) offers the opportunity to leverage machine-to-machine communication (M2M) across secure enterprise networks. Take for example, the determination of opens due to soldering defects on a populated circuit board. This can be communicated upstream where the quantity of a new flux being deposited can be adjusted or the temperature inside the reflow soldering machine can be changed to address those defects.
Connectivity takes relevant information beyond the factory floor as well, allowing manufacturing execution systems (MES) to be integrated with enterprise resource planning (ERP) systems. Real-time data flows can help optimize production, raw material inventories, and the entire operation spanning from suppliers to customers.
The introduction of 5G technology has addressed some of the throughput issues resulting from the vast amounts of data that cannot all reside at the sensor edge. 5G modules have made the addition and management of existing sensors as nodes in a network flexible and easy.
Data Storage: If intelligence is wealth, data is the currency in which it is measured and traded. As today’s manufacturing equipment bristles with sensors, the data they generate must be categorized and organized according to where it might be needed, when it might be needed, and to what end.
As described above, smart sensors can store some data at the edge to better process it locally and enable continuous adjustments, often to a small part of the overall manufacturing process. However, for the purposes of process modeling and development, the vast amounts of data generated over the entire process must be combined. Enterprises must therefore decide what information resides in on-site datacenters and what can be better housed in remote clouds for mining along with ERP information. This eventually drives hardware decisions, such as investments in flash memory that can operate in manufacturing environments.
Data analysis: Data analysis is required to create intelligence for better decision-making. While embedded machine-learning (ML) algorithms make critical real-time decision making at the machine possible, not all decisions have to be made in real time and not all decisions are about controlling a single piece of equipment.
For instance, when developing a new etch recipe in a fab, analyzing all the data from upstream lithography to downstream wafer-level critical dimension (CD) metrology can help reveal process controls for better process modeling. Given enough data to analyze, digital twins — virtual copies of real-world systems — can be developed to allow process simulation and optimization as well as real-time decision making and control. That would not only boost operational efficiencies but speed up R&D and, as in the case of a fab, significantly cut the time and expense toward metrology and testing.
Technology blocks for Industry 4.0, therefore, not only include hardware like sensors but also software, including ML and artificial intelligence (AI) that can turn data into insight and insight into decision making for rapid process and operations adjustments.
Adding value via feedback loops
The consumer electronics OEM value chain has over the years been optimized such that contract manufacturing is part of the production network to deliver high-volume sub-assemblies from lower cost bases. Shorter product life cycles, increasing wages, higher competition, and the pandemic have put pressure on these networks.
Digital technologies have enabled work-from-home for noncritical staff, advanced solutions like augmented/virtual reality (AR/VR), and remote access to process data have ensured that operations as well as staff training continue without interruption.
Under normal circumstances, closed control loops implemented through smart-sensor-based quality control (QC) reduce waste and increase yields by detecting early any process deviations, conducting root-cause analysis, and implementing automatic correction.
Digital tracing of components and products throughout supply and delivery chains enables efficient part movement, inventory keeping, and warehousing to offer significant cost benefits. Traceability through the product life cycle also checks counterfeiting, defect analysis, and product improvement through component substitutions and design iterations.
Industry 4.0’s feedback loops go beyond sensors as a common thread of information pulls together people and machines; design, manufacturing, and distribution; and OEMs, suppliers, and partners into a circular economy.
Assessing smart readiness
The pandemic pushed the launch in September 2020 of the Global Smart Industry Readiness Index (SIRI) Initiative, a collaboration between the WEF and the Singapore Economic Development Board (EDB). The Initiative aims to offer SIRI as the international standard for Industry 4.0 benchmarking and transformation.
SIRI comprises a framework and tools to help start, scale, and sustain digital transformation. The framework consists of three layers that together harness the full potential of Industry 4.0.
- • The Process layer covers integration of operations, supply chain, and product lifecycle.
- • The Technology area includes automation, connectivity, and intelligence on the shop floor, in the facility, and at the enterprise level.
- • Finally, the Organization layer deals with talent readiness and structure and management to assess workforce development, leadership, inter- and intra-company collaboration, and strategy and governance.
The Manufacturing Transformation Insights 2022 report finds, unsurprisingly, that industries like semiconductors and electronics lead other sectors in terms of digitization. It identifies one key insight from “best-in-class” achievement on the SIRI: successful companies have focused significantly on factory connectivity to better leverage data for generating insights and taking real-time decisions — the “smarts” in smart factories.
Getting smarter with Arrow
The goal of Industry 4.0 is to implement automatic, intelligent, real-time control on processes, which relies on the availability of continuous process data. The hardware and the software are the tools that ensure data generation, its safekeeping, its analysis to arrive at decisions, and the enforcement of the decision through process control.
Arrow’s vast ecosystem of partners in Enterprise Computing Solutions, among the largest portfolios of electronic components, and Arrow Intelligent Solutions, including design services, can help build out smart facility resources quickly and efficiently.