3D CAD Package Tips

The latest version of NX software includes an expansion of the Digital Innovation Platform that has been enhanced with machine learning (ML) and artificial intelligence (AI) capabilities. These new features can predict next steps and update the user interface to help users more efficiently use software to increase productivity.

The ability to automatically adapt the user interface to meet the needs of different types of users across multiple departments can result in higher adoption rates, leading to a higher-quality computer-aided technology (CAx) system and the creation of a more robust digital twin.

Machine learning (ML) is increasingly being leveraged in the product design process for a competitive advantage. ML can deliver valuable business insights quickly and efficiently, and it has the power to process, analyze, and learn from large volumes of data. AI and ML can also be used to monitor the actions of the user, their success and failures, to dynamically determine how to serve up the right NX commands and or modify the interface and leverage learned UI usage knowledge for CAx environment personalization.

“There’s always been a capability-usability tradeoff with CAD applications. The more expansive it gets, the more difficult it is to use and master,” said Chad Jackson, Chief Analyst at Lifecycle Insights. “The Adaptive UI in NX, however, circumvents that issue. It guides users, new and old, to the right function at the right time. Many will benefit.”

The Siemens Digital Innovation Platform is continually expanding to enable customers to create the most comprehensive digital twin of a product, the production environment and of the performance of the product. Integrating ML and AI into NX software offers benefits of speed, power, efficiency and intelligence through learning, without having to explicitly program these characteristics. Customers can enhance the design process and reduce time to market.

Siemens PLM Software
www.siemens.com/plm

Automated vehicles (AVs) will disrupt the way mechanical engineers work with their peers and how they share design information. Here’s what you can expect.

Jean Thilmany, Senior Editor

Tomorrow’s vehicles will be electrified, connected, and will include many automated functions, including the capability to navigate roads on their own—without a driver’s intervention.

In other words, they’ll be incredibly complex systems comprised of mechatronics, sensors, radar, automated functions, and many other features. And, like other incredibly complex systems introduced within recent years, they’ll disrupt the way mechanical engineers work with their peers and share design information.

The age of the automated vehicle (AV), also known as the driverless car, is upon us, says Scott Shogan, connected and automated vehicles market leader at WSP, an engineering firm.

Government officials and industry leaders expect AVs will begin to make their way to U.S. streets within the next decade, says Shogan, who follows AV trends on the local, national, and international levels.

AVs probably won’t be the personal, driverless cars you may be imagining—instead, they’re likely to take the form of shared-ride cars or small buses that pick you up at your home. The driverless vehicle will shuttle you from your door to the nearest automated-bus stop or train station to continue you on your driverless commute, says Shogan.

While AVs may not yet be on the scene, connected vehicles—AVs cousins—are increasingly common today. They include features like advanced driver’s assistance systems with lane centering, adaptive cruise control, and swerve detection, notably seen on Tesla vehicles.

None of these features let the car drive itself, but they do help a driver avoid crashes, Shogan says.

The move toward electric is already underway. GM has announced it will launch AV taxies next year at three sites, Shogan says. “What they’ll look like we don’t know. But it shows how quickly things are moving.”

Volvo announced it will stop making internal-combustion-engine vehicles in the next few years, Shogan says.

Already there are around 350 types of electrical vehicles on the market, O’Brien says. “There has been a huge increase in the complexity of these electrical systems,” he adds.

Future electric vehicles will use sensors to gather instant information about what’s going on in the world around the vehicle to help automated systems make split-second decisions.

They’ll also include 40 percent more hardware than today’s vehicles and have safety, security, and power demands not seen in present-day automotive needs.

Systems of design

As you might expect, automated vehicle design is a different ballgame as compared to the way non-automated and non-connected vehicles are designed today. Whether connected or automated, these vehicles will likely be built on an electric platform, Shogan says.

Engineering AVs and other new, complex machines like collaborative robots require close collaboration among engineers who must take a “systems engineering” approach to design, validation, testing and prototyping, says Martin O’Brien vice-president of the integrated electrical systems division at Mentor, which makes software for electronic design.

Like cobots, AVs will consist of electrical, mechanical, and software systems as well as sensors that provide feedback about the world around the vehicle, he says.

While CAD has a place in automated vehicle (AV) design, CAD models are part of all the models that make up the larger design system, O’Brien says. The CAD models become part of a model-based system engineering process,

Which is a method of systems engineering based on creating and exchanging models rather that documents.

All the systems within complex machines like AVs need to be designed to work together, O’Brien says. So software and mechatronic models are shared for a systems-engineering approach to design.

A systems-modeling approach allows engineers to bring together all the knowledge from the multiple domains into one location, he says.

“Systems that used to be tested independently of one another are now so tightly integrated that we have to engage all the systems and test them simultaneously,” he says. “The electrical system has to be designed in the context of the software system and the entire 3-D system.”

Recognizing this, the makers of engineering software are increasingly creating software that can help carry out the design of AVs across the lines of engineering—electrical, mechanical, and software.

The software depicts how all these functions will work together in a real-life operating machine. When the digital systems are all running at the same time and in intended manner, developers often refer to the product mock-up as a digital twin.

SimCenter 3D, from Siemens PLM Software, for example, allows engineers to create models to connect their designs and to carry out 1D simulation, test, and data management functions on those designs, says Dave Taylor, Siemens vice president of global marketing.

While the solution is tied to Siemens NX CAD software on the mechanical-creation side, it can import geometry from any CAD source and prepare analysis models for a range of multiphysics simulations including finite element, boundary element, computational fluid dynamics, and multi-body dynamics, Taylor adds.

Usually a company will use its product lifecycle management system to coordinate efforts across various applications, he adds. To take the Siemens example, Teamcenter PLM, from Siemens, can be used to coordinate mechanical designs created with Siemens’ NX software, and electrical designs created with Capital electrical design software from Mentor, a Siemens company, O’Brien says.

Once it meets everyone’s specifications, the model of the entire system—the entire AV in this case—becomes the digital twin, used to virtually test the AV. Because the digital twin simulates exactly how a machine will function, engineers are able to find and fix design problems before the expensive products are produced.

This is where the digital twin comes in. The digital twin links those models to focus on one aspect of a product or the entire product.

While the CAD model is a part of a digital twin, so are the models that make up the electrical and software systems. Sensor information is incorporated into the digital twin as well, O’Brien says.

“Developing a functioning twin is a systems integration task,” he adds. “Model-based engineering is central to successful outcomes to integrate functions, devices and signals into the platform.”

Simcenter allows engineers to share their model across domains to create a system-wide simulation of a product to be used for a digital twin, Taylor says.

A digital twin is increasingly used for testing today because building a prototype to test all the functions of sophisticated machines like an AV or a collaborative robot is often just not possible, O’Brien says.

“Multi-physics simulation is critical for autonomous vehicles, where the digital twin can drive billions of virtual miles and our solutions can predict exactly what’s going to happen in the real world,” Hemmelgarn adds.

Some parts of the AV won’t require a systems-engineering approach.

The way CAD works in the automotive industry today will still be part of the design process when other aspects like electric and software don’t come into play, for example, for the features found inside the AV.

Of course, the interior of the vehicle will look quite different than it does today, as there won’t be a need for the steering wheel, or really, the driver’s seat. To save space within the vehicle’s compartment, riders might sit facing each other, or they may not sit at all, Shogan says.

But a seat is still a seat. And mechanical engineers will be called upon to use their CAD systems for seat design.

Perhaps the interiors of some AVs will look like the trains in airports that ferry passengers between terminals: they’ll include a few seats around the perimeter, though most of the AV’s occupants will remain standing for their short ride.

On the right track

But testing and validating AV design goes beyond the digital twin, of course. Testing the AV needs to happen with a real-life vehicle.

Safe validation must include a structured combination of three methods: testing the digital twin, testing the vehicle on a controlled track, and testing it on the actual road, says Mark Chaput, vice president of construction and infrastructure development at the American Center for Mobility (ACM) in Ypsilanti Township, Mich.

The Michigan center is an AV proving ground where the technology can be tested safely. The U.S. DOT designated 10 of these sites in 2017. After a testing period, officials from those sites will their share best practices to safely test and operate AVs. That information will enable participants and the public to learn about AVs and will accelerate the pace of safe deployment, Chaput says.

ACM is an open track that companies come to independently to test their AVs. AVs can operate at the proving grounds much as they would on the street, so engineers can get sensor feedback and, of course, perfect their vehicles.

The proving grounds are necessary because these prototypes need to be tested within a closed environment.

“You won’t be able to develop this technology in a traditional way,” Chapaut says. “Vehicle makers need to test perceptions of the world around the AVs and traditional automotive proving grounds aren’t equipped for that.”

The test vehicles will need to be tested across millions of miles to verify all its technological functions, he adds. “AV makers, for example, wonder how to integrate Lidar and radar and the many sensors needed to automatically see the road into their AVs,” he says.

Lidar allows AVs to calculate the distance to an object. It measures the distance by illuminating a target with pulsed laser light and measuring the reflected pulses with a sensor.

In May, ACM brought in Prescan, part of the SimCenter suite of Siemens simulation and test solutions. The program is used to physically and virtually test and validate AVs and connected vehicles.

It produces physics-based simulation of raw sensor data of the potential driving scenarios and traffic situations in which AVs could drive themselves. By running these kinds of tests and simulations, developers can better understand how to position Lidar and radar on their vehicles, as well as improve upon many other aspects of design.

When the AV prototype is finally ready, it can be tested in real life, on the ACM roads, which include stoplights, curbs—essentially everything found while driving on a road, including bikes and pedestrians, Chaput says.

When driveless vehicles finally do become a common part of the landscape, you’ll know the time, commitment, collaborations, and the many tests and validation scenarios that went into their safe creation.

Siemens PLM Software
www.plm.automation.siemens.com