A significant trend in the last decade in manufacturing and engineering can be summed up in a simple phrase—more digital. From CAD designs to sophisticated simulations used to verify their functionality, products as complex as airplanes and automobiles live their early conceptual lives in a digital world. Increasingly, the factory systems used to create them are doing likewise.
Digital manufacturing’s rise will become ever more prevalent as computing gets more and more powerful. New data sources, such as from the Industrial Internet of Things and 3D scanners, allow higher fidelity and accuracy in factory simulations.
Easy-to-use factory simulations are a key to wider adoption.
“We believe strongly that engineers should be spending their time solving important problems for their organization and not being programmers,” said Robert Kranz, director and general manager for the Arena business of Rockwell Automation.
Arena was one of the pioneers in discrete event simulation applied to manufacturing. Discrete event modeling depicts the behavior of a complex system as a series of well-defined and ordered events.
“We use a flow chart technique that looks a lot like [popular] flow charting tools, so you build up to the workflow and process, visually adding in information with mouse clicks that avoids programming,” Kranz said, explaining the key to making Arena’s simulations easy to create and use. It can display results in either a 2D or 3D format, employing a gaming engine to produce its 3D graphics.
Manufacturing simulation tools are well suited for companies that are struggling with at least one of three challenges, he said: A high degree of uncertainty or variability; constraints on resources, or highly complex manufacturing systems.
These SIM tools can also be used to map out and simulate business processes, as well, by using the same flow-chart technique. “A variety of industries use our software, with manufacturing clearly number one,” Kranz said, noting that more than 52,000 college students worldwide have downloaded the software.
Supply-chain analysis and human resource planning are common uses of the Arena technology, he said. Automotive OEMs and small-part manufacturers at the sub-tier level use it, as well, to understand and modify their processes.
The source of the data is an extensive library of common data objects—think a conveyor belt or packaging machine selectable from a menu. Users can also create customized templates.
When asked about accuracy and potential error, Kranz re-directed the question into a discussion of how exactly these kinds of simulations offer value. “[They] really provide value by characterizing and understanding the variability that exists in a system,” he said. “For example, we do not model the thermodynamics or physical properties of the system. If we are modeling how a part is created through a welding process, we look at the statistical variability around that step and use that to understand the variability in the entire process.”
For maximum utility, Kranz encourages users to build a functional specification of the process to define the objective and describe the system and rules of operation, with special consideration for the data required. “Once you do that, creating a simulation comes pretty quickly,” he said.
FlexSim, an independent company offering a software package of the same name offers another simulation tool for factory simulation.
“Our emphasis has been from the start on 3D simulations with realistic and stunning graphics to solve modern manufacturing problems,” said Markus Cueva, educational program director for the 18-year old firm. “Manufacturing is our biggest industry, and the typical users are people with backgrounds and education in industrial and systems engineering, with many holding degrees in that field. Many also are coming from lean and agile mindsets and are used to thinking about how to improve their operations.”
Manufacturing systems that are replicable with certain elements of randomness get the most value from being studied by simulation, he said.
“Variation can come from human elements or a machine,” Cueva said. “Many companies using this are automakers. Sometimes we get into aerospace depending on what they are trying to figure out.”
Aerospace uses simulation more for studying logistics and MRO facilities, he said.
About half of the companies that bought FlexSim use it to simulate completely new plants. The other half use it to study the impact of plant upgrades.
New and existing plants game
“New plants are the most obvious use case, where you can derive ideas and expectations of what you want a new plant to accomplish,” Cueva said. “However, it is useful in existing plants where customers cannot quite identify existing inefficiencies or bottlenecks. It even lends itself toward job shops, trying to make the best use of the resources they have available.”
Data sources are, of course, an important element of any simulation.
“Our simulation can drop CAD layouts of the facility directly into it, using a variety of existing formats,” such as Autocad, STEP, or NX, he said.
Today, it is much easier than it was even five years ago to collect data today from machines, IIoT, embedded sensors and other sources. Still, getting data and creating models are significant challenges.
‘It takes too long to make models’
“Simulation, as much as it has been used, has suffered from time issues,” Cueva said. “It takes too long to make models.”
In his experience, building and exercising a plant simulation is often scheduled at the end of a development project, often to validate what has already been designed.
“Look for our company to create less complex user interfaces with more pre-built options in industry specific modules,” Cueva said.
Visual Components, another independent digital manufacturing simulation firm, emphasizes 3D simulations with reusable simulation components.
“We like to call it predictive manufacturing,” Visual Components North America CEO Robert Axtman said. “We can predict equipment, the throughput, cycle times and cost of the whole manufacturing process regardless of product, geography or industry.”
Paper sketches still in play
Engineers still today often analyze manufacturing systems with rough sketches on paper and data linked in Excel spreadsheets, he said.
Visual Components’ technical approach includes discrete event or material flow with robotics motion, including off-line programming (OLP), resource and process planning and PLC validation all in one product, Axtman said. “Of course, others can do the same thing, but instead of a modular approach where you need to incorporate a number of products to do that, we do it all in one.”
When asked about who uses these simulations, he agreed with others that his company’s “typical user” defies easy categorization. The users range from mom-and-pop job shops up to automotive OEMs. But they all need the same thing—to quickly and easily build a simulation model.
Recognizing this, Visual Components, like others, created an extensive library of pre-built and validated manufacturing components, especially in robotics.
“We have over 1900 different components, including 1200 robots, in our library—from all of the major providers, including grippers and fixtures. We put new ones in as they are offered,” Axtman said, noting that all are part of the standard license. “It is not something you have to pay extra for.”
External data, such as CAD files or PLC interfaces in all popular formats, can be imported, as well as performance data from MES or PLM systems.
“Our software is easy to customize and easy to use. Everyone from manufacturing engineers to front-line sales can build and present manufacturing layouts,” he said. New components can be added that might be unique to each customer through an API.
Sales teams, business managers helped
Axtman also stressed that manufacturing simulations go beyond helping manufacturing alone. Sales and marketing professionals, business managers and companies large and small would find value in an easy-to-use digital manufacturing solution.
The Tecnomatix brand of digital manufacturing software has found itself in a unique position, compared with its beginnings as a tool to help manufacturers set up robots with OLP in the mid-1980s.
After a series of acquisitions, it is now part of Siemens’ product life-cycle management (PLM) software business. “This means our group has responsibility for a broad range of products, some that are included in Tecnomatix, others that are included in the Teamcenter PLM software portfolio, such as manufacturing planning,” said Mike Rouman, senior marketing manager for Siemens PLM Software.
Siemens PLM’s discrete event simulation tool under the Tecnomatix portfolio is called Plant Simulation. Like other solutions, it uses statistics and flow charts to create a “whole plant simulation.” Plant Simulation uses an object-oriented architecture and modeling capabilities, including advanced control mechanisms, the company said.
The interface is modeled on Microsoft Windows standards to make it easy to use and understand. Tools for automatic optimization and 3D visualization of simulation models are also available.
Most customers use a mixture of 2D and 3D simulations, typically on their desktop, Rouman said. “It does not require a lot of computing resources and takes a matter of minutes to build a plant simulation, using drag and drop to place objects that will give you some meaningful results.”
Many uses possible
Tecnomatix can be used for not only discrete event simulation but also simulating both subtractive and additive manufacturing environments.
“For example, we can simulate workcells or workstations using a cyclic event evaluator,” he said. Instead of estimating when an event is triggered in a time-stream, the engineer can specify an event that triggers the required action, providing a more accurate simulation. “If I want to simulate and see my virtual robot cell operating as a function of a signal that is coming from a controller, say a PLC, and I want to check back that it works correctly, I have to simulate based on events,” he said. “We can use data from these types of simulations, as well as data collected from actual operations, to feed ever more realistic plant simulations.”
Source data can come from IIoT, MES or PLM inputs like Teamcenter, he, like others, noted.
Video games create new expectations
Siemens will in the future address the continuing challenges of reducing the level of effort needed to get a good result.
“We are challenged with the perception that simulation should be easy,” Rouman said. In part, this is fed by how easy yet realistic today’s computer video games are. This perception produces an underestimate of how difficult it is to produce statistically realistic manufacturing solutions, despite the great strides in usability companies like Siemens and others have made.
Dassault Systèmes, another large PLM supplier, also offers a digital manufacturing suite in its DELMIA brand. DELMIA digital manufacturing applications deliver simulations within ergonomics, machining, process planning and simulation, manufacturing management, manufacturing resources and robotics.
“There is a lot you can achieve with digital manufacturing,” said Patrick Michel, marketing VP of DELMIA Digital Manufacturing for the French firm. “We derive a lot of value tying digital manufacturing closely to the engineering space. We are well known for CAD products like CATIA and SolidWorks, and we have derived a lot of value by tying our digital manufacturing offerings closely to the engineering side of our business. This is especially true of integrating changes, capturing them before, for example, tooling is built, or uncovering unbuildable conditions and rectifying it before substantial cost is incurred.”
The total package includes discrete event simulations, kinematics and ergonomics, allowing manufacturing engineers to review many facets of a factory system.
Tool-specific online community built
Dassault is taking advantage of social media by building DELMIA-specific communities online.
Four in particular mirror how digital manufacturing is used. The first is process planning and simulation, and it is aimed at bigger picture questions around manufacturing systems. The second is robotics, dedicated to the needs of robot programmers. The third is fabrication, which addresses CAM and CNC machining and now includes additive manufacturing. The fourth is ergonomics.
An interesting observation Michel made is that even if a simulation is used to build a factory, once built, it soon deviates from the ideal.
There are a number of reasons, Michel said, including lack of priority in collecting data once the factory is built, along with added expenses.
“We were once confined to adding value at the beginning of a program when things were created from scratch,” he said.
In part to rectify getting relevant data later for simulations, DELMIA acquired two MES systems, Velocity Intercim and Apriso. “The idea was to connect and leverage that asset we had in digital manufacturing to the actual operations on the shop floor—bi-directionally,” Michel said. “We are working toward the ability to take any factory that is running Apriso, take it off line and run simulations to plan how to improve it. The starting point then is what is happening on the shop floor right now, using actual data from Apriso.”
This notion of actual data versus original specification extends to the physical layout, as well.
Most models used to study and improve existing systems include incorporating data from 3D scanners and/or photographic data, Michel noted.
“I see digital manufacturing moving out of a proverbial dark cave, with a few users doing some super things for a larger audience,” he said. “It has become more visible to the larger organization than the manufacturing professionals that use it.”