Issue link: https://iconnect007.uberflip.com/i/1091269
56 PCB007 MAGAZINE I MARCH 2019 in productivity, efficien- cy, savings, yields, au- tomation, enabled trace- ability, compliance, and reduced risk of errors and rework. All of these items are crucial factors when manufacturing printed circuits. In a recent publication on smart factories, Deloitte University Press [2] stated the following: "The smart factory represents a leap for- ward from more traditional automation to a fully connected and flexible system—one that can use a constant stream of data from con- nected operations and production systems to learn and adapt to new demands. A true smart factory can integrate data from system-wide physical, operational, and human assets to drive manufacturing, maintenance, inventory tracking, digitization of operations through the digital twin, and other types of activities across the entire manufacturing network. The result can be a more efficient and agile system, less production downtime, and a greater ability to predict and adjust to changes in the facility or broader network, possibly leading to better po- sitioning in the competitive marketplace." Not Necessarily Smart A factory itself cannot be smart; it needs smart technology to support. For a smart fac- tory to act, it needs data. For this, we need smart technologies that make the appropriate data available so that the machines can do their job. No data input or output. IoT, API, AI, and Industry 4.0 allow large amounts of man- ufacturing data to be collected, transferred, read, and interpreted through multiple sources and cloud-based, secure storage without any human involvement. However, machines tend to be bad interpreters. For machines to interpret and make use of data and enable computer-talking-to-comput- er, they all need to speak the same language. The supply chain has evolved from some points of contact to a complex chain of design- ers, OEMs, assemblers, brokers, distributors, factories, and subcon- tractors all sending data back and forward, doing manual labor trying to interpret the data. If we cannot interpret the data correctly, how can we expect the machines to do it? Still, Rome was not built in a day, and progress is happening as we speak. We saw the need for this common lan- guage for specifying printed circuit fabrication data, which is why we initiated CircuitData. In Need of Robots Again, smart factories are characterized by adaptability, resource efficiency, and ergonom- ics as well as the integration of customers and partners in business and value processes. The goal of a smart factory is to optimize the con- cept generation and transform the production and the production transaction into a more ef- ficient process. It is a subset that employs com- puter control and high levels of adaptability. When visiting smart factories, the one thing that instantly strikes you is how little human involvement there is at the plant [3] . But smart factories do not operate by them- selves; they still need the workforce, which is where robots enter the picture. However, this is not a new asset to be added to the manufactur- ing floor. Robots and humans have been work- ing side by side for many years. What's new about the robots of today is they are smarter, more independent, and can perform several tasks without stopping the workflow and the need for reprogramming. Again, AI is the key. Connecting the Dots With Data We all know that the future is all about da- ta. Imagine how things are today—a supply chain built partly on paper and digitally. The tracking and controlling process are time-de- manding, and it's difficult to have full control and a broad overview. In PCB manufacturing plants, large amounts of design and produc- tion data are held in many places and differ- ent formats. Jan Pedersen John Steiner