In 2007, Klaus Schwab told an assembly of the World Economic Forum that the world was now entering what he called the “Fourth Industrial Revolution” — a new period of manufacturing technology that will be fully upon us in the next ten years. At its core will be cyber-physical systems made up of processors, software, sensors and communication technologies.
If the Digital Revolution of the second half of the twentieth century brought a shift from analog and mechanical technology to digital technology, “Industry 4.0” describes the linking of production and automation technology with the Internet, optimizing production in real-time. Powered by the convergence of low-cost storage, ubiquitous sensors, powerful artificial intelligence and analytics, plus widespread adoption of Software as a Service (SaaS) and Cloud-based computer systems, one can think of Industry 4.0 as the interaction of the real and virtual worlds — with huge impacts for product engineering and manufacturing.
The first hints of the power of Industry 4.0 can be seen in the increased automation of various industrial processes in fully digital factories. The first fully digital factories have been implemented by companies who have integrated software platforms that link all aspects of manufacturing, from initial product design to factory layout and manufacturing process optimization to customer feedback after delivery. This integration ties together each aspect of the manufacturing process, including CAD modeling and visualization tools, design for manufacturability (DFM) analysis software, computer-aided manufacturing (CAM), computer-integrated manufacturing (CIM) and shop floor data collection systems.
These software systems do more than trade data. They eliminate much of the human intervention that goes with managing them. This next generation of automation, utilizing big data, analytics and artificial intelligence, is one of the most important drivers behind digital manufacturing and Industry 4.0.
Industry 4.0 in the Real World, Or Eliminating $60K/Hour in Scrap or Rework
Consider this example of an electronics manufacturer:
Printed circuit board assemblies (PCBAs) are manufactured with 1,000s of components using 100s of different component part numbers. Using surface mount technology (SMT) machines, a single “reel” of components that are fed at high speed into the SMT machine can have hundreds or even thousands of components on a single reel — and modern SMT machines can place more than 100,000 components per hour. If an incorrect component is loaded onto the machine, thousands of PCBAs will be incorrectly built.
As in any manufacturing industry, reducing or eliminating waste and rework is an important focus, as significant costs can result from having to rework the PCBAs or recall goods after they have been shipped. In electronics manufacturing in particular, the speed of manufacturing processes can lead to extraordinary costs: a typical line might be producing one million units per month, with a cost of $30, and the line runs 20 hours per day, 6 days per week. At this speed, if the wrong component is place by the SMT equipment (or if it’s orientation is incorrect), a factory is generating $60k per hour of rework or scrap.
Various systems have existed for a number of years to prevent loading the wrong components and offer some traceability in case of recalls. However, most legacy systems still depend on people in the stockroom to print and attach barcode labels onto component reels, and have a production operator then scan that label when reels are loaded onto SMT pick and place equipment.
The limitations of these human-dependent processes are displayed when examining recalls issued by the FDA in 2016. The production date ranges in each recall typically span several months — and can stretch to years! Clearly there is a need for increased accuracy in traceability, as well as improved error-proofing and controls.
One company recently brought together current innovations in SMT equipment, the implementation of 2D barcodes etched directly onto the printed circuit board, along with an advanced MES to ‘error proof’ PCBA and system level traceability. With this implementation, the SMT line is in direct communication with the PLM system, and will prevent the operator from loading an incorrect component (reel) part number.
More importantly, the scanning of individual serial numbers on each PCB enables a direct correlation between the PCBA serial number and the date codes and rev levels for each component used. Machine to IT communication enables the change in component and the new reel to be recorded in the manufacturing execution system. Moreover, since the MES system communicates with the PLM database in real-time, if an engineering change order (ECO) is implemented that updates a component revision or part number, the SMT equipment “knows” this in real-time.
In the event of a component recall, complete traceability information is now available at the PCBA and system serial number level, and is recorded digitally in the MES and electronic traveler. So, it is no longer necessary to recall or quarantine excess products built over multiple days or weeks in order to ensure that all non-conforming products have been recalled — manufacturers can now recall or quarantine ONLY those parts that it knows contain defective components.
This combination of intelligent systems virtually eliminates human error, eliminates the rework and recalls associated with human error and limits the scope of potential recalls when a supplier notifies the manufacturer of bad components, as well as speeding notifications about potential problems. This is just one example of the power of digital manufacturing and Industry 4.0, though it is one with obvious cost and compliance benefits.
The increased velocity of digital manufacturing offers even more benefits: cost savings in the product design process; greater efficiencies for on-demand or low-volume manufacturing; and accelerated design iteration that leads to greater innovation overall for developers.
How to Get Started — And Manage the Risks
For many manufacturers, change equals risk. Disrupting their supply chain or factory lines can cost millions, with untold damage to a brand and a company’s reputation. Given this, many manufacturers are reluctant to embrace far-reaching changes to industrial infrastructures and processes, understandably. However, manufacturers can start down this path by resetting expectations and asking their personnel to think differently.
Take a look at the status quo, examine the deficiencies that you’ve come to accept day to day, and use these as drivers for change. And don’t forget to identify your most digitally savvy supplier; partnering with these suppliers to learn about their tactics and operational improvements can lead directly to improvements in your own operations. Everyone — including executive teams — must recognize that the path to digital manufacturing is going to be different than previous operational changes.
There are no hard and fast rules to Industry 4.0 — but there is great promise in this coming wave of change.
Srivats Ramaswami is CTO at 42Q.