Intelligent technology based on data, change the face of production

ASTM International is developing fundamental standards that will help make Industry 4.0 a reality.

Think for a moment about how the brain approaches a new task and learns to cope with it.

Suppose you never baked a cake, but you want to make it for a special occasion. What are you doing?

Today, the first step may be to go online and watch a video with instructions. Each time you watch one of them, your brain registers steps in this process and begins to understand how they fit together. Watch enough videos, and you will assimilate this data and be able to act independently.

The underlying premise of Industry 4.0 is very similar, only on an exponentially larger and unimaginably faster scale. Imagine that every hour, thousands of bakers make millions of cakes. Consider each step in the recipe of each baker as a data point. As if did you cope with such a huge accumulation of information and, more importantly, how would you use it to improve the process? These issues are especially relevant in the world of additive manufacturing (AM). This extremely time-consuming process creates three-dimensional objects by connecting successive layers of material, one on top of the other, by directly converting three-dimensional files of computer-aided design, or CAD, which contain all the necessary design parameters.

The following is an overview of the work of the ASTM International Additive Manufacturing Technology Committee (F42), which is at the forefront of developing globally relevant AM standards that are key to the development of Industry 4.0. Since its inception in 2009 over the past year, the committee has developed more than 25 critical AM standards, and the recent creation of a subcommittee (F42.08), specifically focused on AM data, will take Industry 4.0 even further.

What is Industry 4.0?

Industry 4.0 (abbreviated I4. 0) is a term used to define the process of integrating digital technology into physical production. Essentially I4. 0 includes the collection of a huge amount of data generated by modern automated production systems; Evaluation of data to identify patterns that may reveal an understanding of problems or better ways to accomplish tasks; and the inclusion of these decisions in the production process both in the framework of real-time decision-making and in the framework of a long-term analysis of quality control.

“Industry 4.0 is a digital transformation,” said Alex Liu, Program Manager for Additive Manufacturing at ASTM Asia Pacific. "This is the place where cyber-infrastructures, such as artificial intelligence [AI] and machine learning [MO], collaborate and integrate with physical infrastructures, such as additive manufacturing and robotics automation. This is a digitization of production. "

"In fact, Industry 4.0 is a combination of machines, work cells and production halls with an information infrastructure in which decisions are made," says David Rosen, professor of mechanical engineering at the Georgia Institute of Technology and chairman of the AM Subcommittee for Design (F42.04).

The digital ecosystem that collects this data is based on the so-called Industrial Internet of Things, or IIoT. IIoT refers to machines and systems that connected and interact with each other over the Internet.

“AM is just another type of machine that will be connected,” says Rosen. "In the case of AM, this technology is closely connected with other components of Industry 4.0, such as robotics and automation, big data, cybersecurity and artificial intelligence. The merger of all these technologies provides the greatest opportunity, which is to extract new ideas and discover patterns," says Mahdi Jamshidinia, Ph.D., is the director of the ASTM additive manufacturing research project. It is this connectivity, as well as Artificial Intelligence algorithms that recognize patterns (and anomalies) in the data stream, that can bring transformative benefits a AM industry.

“For example, machine learning is used to quickly select the best AM source materials and can help predict the functionality of AM parts on based on several design parameters, ”explains Mohsen Seyfi, director of ASTM's Global Additive Manufacturing Programs.“ Artificial intelligence can also monitor the AM manufacturing process in real time and analyze the root causes of manufacturing problems. ”

Seyfe also points out that Artificial Intelligence can be incorporated into AM systems through the IoT. “For example, you can place a barcode or label on each AM part and use the I&T to track its performance and quality. Then, the data collected from these sensors, could be analyzed to identify trends or weed out other useful information. "

Additive Manufacturing and I4.0

Fundamental principles 14.0 are applicable in most modern production environments, where digital sensors and controls are used to some degree. However am is especially instructive "laboratory" to verify and clarify these principles, since the amount collected the data, even for a short period of time, is staggering.

"AM generates a tremendous amount of data - from AM feedstocks, design, modeling and production processes before post-processing, control, testing and, in ultimately, the performance of manufactured parts, "said Jamshidinia.

is it huge? According to Rosen of Georgia Tech, every minute a machine can work turn into gigabytes of sensory data.

One of the sources of all this data is a painstaking, layered process that defines additive manufacturing (other data sources include product life cycle and value chain activities). Each individual object created using AM, may contain thousands of layers. Sensors capture data for each layer by analyzing and evaluating whether they were handled correctly.

These sensors can also be connected to IIoT, providing access to computing power, necessary for the effective use of all collected information. "Exist a great opportunity to use this data to accelerate the development of the AM industry through

Industry 4.0 applications such as artificial intelligence and large analytics data, ”says Safie.

Vista I4. 0 in the context of additive manufacturing - it is constantly growing, more and more accurate automation of the processes of analysis and evaluation of accumulated data. "Full inspection of each individual The details are labor intensive and expensive. But if there is enough full-time critical datasets machine learning can greatly simplify the necessary verification process, identifying suspicious areas of each part and while ensuring production of quality components, "says Liu.

Data management.

Given the importance of data at AM, ASTM Additive Manufacturing Excellence Center International (AM CoE) in collaboration with America Makes organized the AM Data workshop Management and Schema Workshop last December. This event gathered ver 90 experts from government, industry and academia to cover the latest AM data management processes and data-enabled applications, as well as for discussion gaps, problems and potential solutions for AM data. One of the outcomes of the seminar was creation of a separate subcommittee according to AM (F42.08). "As progress progresses, standards, covering compilation, organization and distribution of AM datasets will fall under the auspices of this new subcommittee, "says Jamshidinia.

Managing this flow of information is fraught with a number of problems. Firstly, it’s itself the nature of the data itself. Member, Subcommittee on Data and Scientist, National Institute of Standards and US Technology (NIST) Yan Liu explains: "Multimodal data is generated over AM product development life cycle, including 3D models, on-site measurements and post-production data inspections. Each of the data types represents the spatial information of the part in different reference systems. "

The proper alignment of these data types within a common frame of reference is called data logging is an approach that allows you to correlate and interpret the relationship between AM process, the properties of the materials used and the object itself. Deeper study AM technology shows the complexity of the process For example, there are currently at least seven different AM categories established by document "standard terminology for additive manufacturing - general principles - terminology "(ISO / ASTM52900).

Finally, an almost unlimited range of products can be manufactured using AM, with all the variations in size that such a blank canvas implies. Be it components rocket engines, orthopedic implants, detailed architectural models or aftermarket parts for vintage cars, AM allows you to produce objects that just can't be done in any other way.

Given all these factors, it becomes easier to assess the complexity of matching all the data, associated with AM processes, materials, and specific details. "Currently large part of the data received from AM activities is collected and stored in a special form (e.g. in PDFs or home spreadsheets), which prevents their use for decision making, ”says NIST Yan Lu, who is also a deputy chairman of the subcommittee for data.

Alex Kitt, chairman of the data subcommittee and product manager for EWI, reiterates Lu's comments on inefficient and inconsistent data collection. "Often the person taking the measurement brings the USB drive to a data specialist who then enters the data into the system, "Kitt says." This prevents any data from being received, which are not critical. "

Lou and her colleagues believe that standards are needed to address key issues such as ways to safely collect, store and process data, as well as determine which formats should be used to exchange data between different stakeholders, related to product life cycle and supply chain management. "Standard AM information models and data change formats are necessary for sharing and complete using data from various stakeholders, "she says.

As production becomes digital and processes become more modular, additive manufacturing requires the exchange of information between numerous stakeholders in the production chain, from design to end users. “It requires robust cybersecurity so that data can be safely be transmitted in a cloud computing environment, "says Safie.

First steps.

The first task of the data subcommittee is to create a single data dictionary (CDD) for the additive production. Lou says: "NIST has been leading the development of CDDs since the fall of 2018 through a special a working group with around 50 experts from industry, government and academic circles.

"Stakeholders represented in this group include NIST, EWI, Penn State University, Granta MI (developer of materials management information systems) and various government groups.

According to Lou, the draft CDD version 1.0 is currently under consideration, and parts of it will be included in ASTM international standards. The next steps are working on a common model production data, data exchange formats and data recording standard AM.

Another priority area for the subcommittee will be AM technologies based on artificial intelligence, including machine learning, deep learning and intellectual data analysis. Seyfe notes that, although rare, AI is already used in additive production.

"For example, the ALCHEMITE ML algorithm was used to develop a nickel-based alloy for direct laser deposition, "he says." In addition, cloud infrastructure was carried out 3D analysis and reconstruction of additively manufactured materials. No standards setting consensus-based methods of using artificial intelligence, organizations find it difficult to replicate, develop or expand these efforts on a significant scale. "

In the meantime, an AM Strategic Data Guide is being developed based on materials from various sectors. "This guidance will be provided to Subcommittee F42.08 and AM community as a reference. This will allow stakeholders identify existing gaps and challenges, and suggest potential future solutions to improve data management and use through timely development relevant standards, "says Jamshidinia.

Inter-committee collaboration is a hallmark of the standards development model ASTM. In the case of the Additive Manufacturing Technology Committee (F42) and its subcommittee according to new data, it is planned to work with other committees dealing with issues AI and IoT / IIoT, including unmanned automatic committees guided industrial vehicles (F45), unmanned aircraft systems (F38), consumer products (F15), and medical and surgical materials and devices (F04).

While autonomous forklifts, unmanned aerial vehicles and intelligent baby monitors provide an opportunity to explore the intersection of data, connectivity and performance, additive manufacturing takes this relationship to a new level. The work of the data subcommittee will undoubtedly contribute to this bottom-up trajectories.

“AM is the first new manufacturing process developed in the data age,” concludes Alex Kitt "However, ensuring that the data is understandable, high quality and usable to use is difficult. A new subcommittee will develop standards, with so that the community can create a data ecosystem in which data can be easily collected, be managed and used. "

Source: https://www.standardizationnews.com