The world of manufacturing, including NC and CNC hardware and software, demands evolutionary changes and transformation. The latest iteration includes cloud, edge, large language models (LLM) and convergence.
Each is deeply rooted in advancements in familiar computing technology edging their way to tool tips. Yet their import must extend from C-suite to shop floor and recoil from shop floor to operational management.
Industry leaders such as San Francisco-based Autodesk Inc. began a recognition-based change a decade ago. “The reality is, all these tools are disconnected, and that’s why we called our product ‘Fusion,’” says Al Whatmough, director, product management, referring to the company’s well-known Fusion 360 CAD system. Because of a lack of large data sets that span these differing verticals, Autodesk sought to get all of the data into one system as opposed to disparate tools that have data in different formats.
That’s been the problem. Unfortunately, there was no large language model available.
“A topic of discussion with every company is about what is most important to a shop,” adds Ben Mund, manager, channel marketing at Mastercam developer CNC Software LLC. “The end result of that discussion usually boils (down) to productivity. Regardless of what software package or machine tool or tooling you use, it’s increasing the productivity of a shop—whether it’s saving time, money or improving the process and output—that’s critical. The concept of digital manufacturing is growing aggressively.”
Mastercam grew during the pandemic “because there was a greater need for information connectivity, a need that was always there but more acutely felt when people were only available to do certain things via digital means,” according to Mund. “And the people who were running the shop floors who could be there physically had an even greater need to have that digital connectivity.”
Volker Nesenhöner, CEO of Open Mind Technologies AG, addressed the company’s software technology development by saying, “It is not enough to just deliver new technologies. CAD/CAM developers must innovate. We are continuously developing new technology to enhance our CAD/CAM software suite. ... We have also strengthened product development and quality assurance—and service and support are increasing in importance.”
Innovation and embedded history run deep in this industry—whether gathering or executing data—but management strategies are widely divergent. Thus, end users, those at the chip-producing end, are rightly concerned about automation.
Thiago Miranda, CNC business development manager for Fagor Automation (part of Spain’s Mondragon Corp.) in Rolling Meadows, Illinois, thinks ease of use and ease of customization simplify machine use and help maintain operator concentration. “Our OEMs build machines and, similar to web development or app making, our highly intuitive HMIelite will provide OEMs an approach to developing new (control system) screens, new types of interfaces and, by using HTML 5, deliver web-based development tools,” Miranda says. “We want the machine builders to not only take advantage of the customizable interface of the control to make it their own, but to be able to both import third-party software and integrate third-party hardware. As a result, managers and end users will have machinery that is easier to use (and) easier to facilitate operator cross training. ... That will address the other side of the picture, which is companies not having the correct personnel to operate machines, which I believe will solve some operational problems.”
More specifically, Open Mind’s Nesenhöner points to early applications pertaining to digital twins, such as collision avoidance. “Now we enable instant real-time bidirectional communication between the machine control and CAM software to significantly improve safety, reliability, speed and process control optimization from CAD/CAM to machine,” which Open Mind calls hyperMILL VIRTUAL Machining. This prevents machine collisions and downtime as well as optimizes highly efficient multi-axis machining processes, according to Nesenhöner, as it has “knowledge of the part model, tooling, the toolpath calculation and collision check engine, and the machine tool model with its physical constraints.”
Fagor Automation is less involved with digital twins as the company embraces a wide variety of OEM machinery. However, Miranda says the company’s dynamic machine control presets “warnings and alarms based on tool location, control of torque, needed tool replacement or other mechanical problems causing stresses in the machine.”
Mastercam’s position, according to Mund is: “Our software powers machine tools like many other CAM packages, but that’s just one part of what a shop needs to do. As any software developer that is invested in helping a company take advantage of digital threads must provide a platform for that shop to do related operations like metrology, robotics—even digital tool management.”
While top-tier manufacturers have aggressively transitioned to expand their enterprise resource planning (ERP) and manufacturing enterprise systems (MES), some smaller shops have lagged behind and haven’t reaped the benefits. But they are now starting to utilize such planning tools and coming into their own.
“The level of potential efficiency improvements and productivity gains are wide and deep for CNC programmers who have access to that real-time data, ranging from refining tool pass strategies to generating cost analysis studies, monitoring machine and operator performance, and troubleshooting issues or improving parts set up and work holding conditions,” Mund says.
AI and Industry 4.0 are more the focus of integrated houses, though their application affects everyone at some level. One example: Toyota Motor Corp. plans to use generative AI to assist vehicle designers. Early in the process, Toyota designers use proprietary algorithms to cut down on hand-drawn iterations and “reconcile design and engineering considerations,” according to the company. This text-to-image generative AI uses optimization theory from computer-aided engineering to optimize, for example, drag (affecting fuel economy) and chassis dimension (affecting handling, packaging and safety).
Open Mind has developed, and is testing, cloud-based solutions while cooperating with German universities to benefit its customer base. Fagor Automation offers a feature to analyze and automatically respond to improper machine operations, delivering optimal results for inexperienced operators or when multiple machines are in operation.
“Our dynamic machine control lives on the CNC, so we are writing new algorithms and adaptive tooling optimization, ensuring tool load matches machining algorithms to maintain predefined machining conditions,” Miranda says.
“Whether it’s cloud or wherever the processing is done, that’s only one component. That’s the delivery mechanism, the data exchange between individualized components that are critical components,” Mund says. “We have, in our industry, a broad variety of partially cloud-based, fully cloud-based, client-side, and server-side communications and web-based components. But that’s just passing the information back and forth. What’s critical is what usable information can be obtained from these components. As with anything, information isn’t valuable unless it is usable. So, what information can then be spread out to other components to make each operation more efficient based upon that information?”
“Innovation is always driven by the speed of communication,” Whatmough says, unlike the era of postal mail or the Pony Express when information was static once delivered. “Manufacturing has worked that way to date. A design gets released, moves into a separate format, and that’s off and dusted. And you have to go all the way back to the beginning and send out a new chunk of data. There’s not really flexibility or any agility.”
Historically, after a design release, data moved into a static format and required backtracking to the beginning to create new instruction sets.
“What we see changing is the ability for change to happen,” Whatmough explains. “Autodesk will openly admit (that) for 30 years we didn’t even think about manufacturing. We just did geometry creation. Now you’ve got workflows where you can machine and measure with the same solution. A user has access to that data all in one place. It’s about connecting the tools and the data together.”
CAD/CAM systems are a basic requirement for creating NC programs on CNC machines, and can also achieve a seamless flow of information and leverage synergies by connecting with other systems and areas, Nesenhöner adds. “We see connectivity as being critical for maximizing manufacturing efficiency. Parts are not only defined by their CAD data but by all the valuable information that is generated in the manufacturing process chain.
“This can include CAM data, NC codes, information on manufacturing parameters and required tools or machining times,” Nesenhöner continues. “The CAM system must provide an interface to all systems that handle manufacturing-relevant data, thereby presenting an ideal way to automatically load CAD data and feed CAM data and NC codes back into the data-handling systems. The CAM system also can play a central role when it comes to tools, requiring it to import data from tool manufacturers and create 3D data from tools for ‘true-to-detail’ simulation. It should also be possible to read tool information from the machine and write data back to it.”
Historical data that was printed and filed, then examined to iterate changes, is no longer acceptable, though entire hierarchies still exist to support such systems. The speed of iteration now drives innovation—another aspect of LLM, such as ChatGPT and other AI networks.
“What we’re really doing is shifting upwards human decision making,” Whatmough asserts. “As we connect design and manufacturing together, we can start thinking about, ‘What’s the modeling you want to accomplish?’” For instance, instead of additively creating and then subtracting, connect additive and subtractive simulation to negatively bend the parts “so that when it warps in the process, it warps closer to what it was,” which would drive production efficiency.
“What this really does is drive us towards automation,” Whatmough continues. “When we’ve connected all the tools together in the cloud you can see higher level intent happening. Perhaps you’ll run an automation and the user says ‘I actually just want to run a prescribed manufacturing method on this part, have it fully calculated heedlessly outside of a client, generate toolpath in an automated way, but done as if a user did it.’” This seems similar to LLM AI instruction. Whatmough explains that “software writes an essay as if a user did it, but you develop a text file and you can start modifying it. In this case it generates toolpath, but you can open it in our full authoring client and a user can make changes to it.”
Looking beyond 2024, Autodesk sees a shift toward higher-order problem solving. “You’ll see opportunities for folks who understand manufacturing now thinking about manufacturing engineering, or folks who are thinking about manufacturing engineering thinking more broadly downstream because the tools are becoming easier to use and more automated.” Whatmough says. “Thoughts will evolve into thinking more about the process and about design intent. The same is happening in manufacturing. You no longer have to think about, ‘Is this tool path going to gouge my part?’ You should be thinking about what operations, what quality checks do I want to inject into my process to get a good quality part—which shifts the user’s thinking. To me, that creates opportunity as operators take a leap into design and manufacturing as their understanding of how something’s made influences design,” Whatmough adds.
Advanced CAD software also enables greater standardization, automation and, ultimately, optimization of the programming of complex components. Nesenhöner points to Open Mind’s hyperMILL system as an example.
“Errors are avoided because strategies and technologies that have been tried and tested can be easily and automatically repeated,” he says. “This frees up programmers to devote themselves to more demanding tasks, and also helps to alleviate the pressures caused by the shortage of skilled workers, as existing manufacturing knowledge is not lost.”
One example: Automatic processes created in hyperMILL that recognize base and side faces, pockets, holes and 3D geometries, then closes the faces to create the body for roughing. A stock and fixtures order is created, and, once the machine and clamping correspond to the physical conditions, the toolpaths are accurately simulated, according to Whatmough.
Open Mind’s automation techniques include scripts with commands, applying more to workflow and process than to specific geometry sets. Some of the script commands can be set to halt automation and enable case-specific user input.
Fagor Automation’s Marketplace will provide machine makers and end users an opportunity to offer custom interfaces and program cycles, creating a new business wherein sales of CNC software add-ons are available.
“This is a completely different approach, as you can create custom interfaces not tied to a specific machine maker,” Miranda says. “Today you need to know specific types of programming languages and have a programmer available to develop new screens or datasets. We are bringing app-like web development to the machine environment, which opens the field beyond CNC programmers to, for instance, gaming or special effects programmers or any skilled thinker.”
In some cases, lab technicians are already using five-axis machining on a patient-specific part that’s used in the medical industry, after scanning a barcode to download a program, according to Miranda. Fagor’s goal, he says, is to ensure OEM machine builders have the flexibility to make machine operation as easy as using a smartphone or a personal computer.
What is clear, management needs a high-level vision, courage, confidence, and an ability to empower and remove obstacles. Driving change requires intricacy of balance and the ability to execute.
“Every change is opportunity, and that opportunity is going to largely come down to where you make your bets,” Whatmough concludes.
Connect With Us
