Deep learning software specialist Intellegens has enhanced its solution for Additive Manufacturing (AM) with the launch of a new Alchemite™ AM Package. The ready-to-go bundle of machine learning software, analysis tools, and implementation services is designed to help AM teams extract value from their data, optimise build parameters and ensure more repeatable AM processes, while greatly reducing the need for testing. The launch of the Alchemite™ AM Package follows Intellegens’ success at the recent AM Innovation Awards, organised by the American Society of Mechanical Engineers (ASME). Intellegens’ Alchemite™ technology is unique in its capability to extract value from sparse and noisy experimental datasets, such as those generated during AM projects. These projects seek to understand the subtle combined effects of material properties, machine parameters, and operating conditions on the consistency and performance of AM parts. Such understanding helps to get parts to market faster, fulfilling the promise of AM for more flexible delivery of stronger, lighter manufactured products. For further information see the IDTechEx report on 3D Printing and Additive Manufacturing 2020-2030: COVID Edition.

The potential of Alchemite™ as a game-changer in AM development was recognised at ASME’s AM Tech Forum event in June, where Intellegens scooped both the Best-in-Class Software and Startup Innovation Awards. Intellegens was selected from over forty organisations, which presented technical solutions, with an expert judging panel making the final decision in combination with attendee votes at the event. Alchemite™ has been validated in a range of customer projects and through Project MEDAL, an AM collaboration with the UK Advanced Manufacturing Research Centre and global aerospace giant Boeing, supported by the National Aerospace Technology Exploitation Program (NATEP). Intellegens is also working in partnership with engineering simulation leader Ansys to embed Alchemite™ within its AM data management solution. The Alchemite™ AM Package builds on the knowledge developed through this work. It delivers the right combination of software and expert advice to ensure that new user organisations get up-and-running quickly as they apply machine learning to accelerate the delivery of AM processes.

The post Intellegens Launches Machine Learning Package for 3D Printing appeared first on Artificial Intelligence.

We have seen machine learning transforming every field that it touches. The field of manufacturing, too, has witnessed many applications of data-driven solutions. Building rockets is unlike any other manufacturing job.

For example, the cost and time it takes to build a decent rocket engine require a multitude of variables. These variables influence how stable the engine is and helps assess other such attributes. A rocket is a combination of more than 1,00,000 individual parts, and electronic equipment joined together meticulously to endure the immense forces of our planet. So, where does AI fit in? Can it overtake the traditional rocket-building methods?

In the next section, we shall look at a few of those solutions and what implications they have:

Finding Stability Of Rocket Engine

A group of researchers from the University of Texas at Austin are developing a machine learning-based framework that blends in scientific computing through a combination of physics modelling and data-driven learning. These ML approaches will then be used to build simulations – or what they call – reduced-order-models or ROM, which can then be used to experiment with different design parameters in a fraction of time.

These reduced-order models help accelerate the designing process and save a tremendous amount of time for the design engineers.

Their techniques are aimed at finding the most stable design for the rocket’s engine. The stability of a rocket’s engine, which must withstand a variety of unforeseen variables during any flight, is a critical design target engineers must be confident they have met before any rocket can get off the ground.

Tuning A Rocket Engine With RL

Of the many things ML could be used for in the case of rockets, their engines seemed to be the crowd favourite. The performance of a rocket boils down to how good the engine is. A team of engineers from Insights Fellow program have chosen a fancier route to improve the engines.

They have employed reinforcement learning (RL) approaches to cut down the time taken for analysis. RL has been applied to create an impactful solution, a control policy for the engines that are as good as what decent control engineers would recommend. This is believed to be several months of trial and error, and brings the rocket to the launchpad much earlier than what it takes for traditional methods.

Using 3D Printing Combined With AI

So far, we have discussed how machine learning is being used to analyse the results and to recommend optimal ones at the post-production of an apparatus. But the engineers at Relativity Space have taken a more ambitious route and have applied ML to monitor the manufacturing of propellant tanks and other large objects. And, they manufacture these large metal bodies using 3D printing techniques!

Apart from Relativity, there are companies like SpaceX, Blue Origin, Rocket Lab that are using 3D printing techniques to print select parts, but the scale at which they are using AI and 3D printing is unprecedented.

Relativity happens to be the first and so far the only company that has blended the advantages of intelligent robotics, software, and proprietary metal 3D printing technology to automate aerospace manufacturing.

These printers use a different printing technique, in which a laser welds together layers of ultra-fine stainless steel dust.

In a recent interview with Wired, Relativity’s co-founder spoke about how they are using AI to improve their manufacturing process. He explained how artificial intelligence tells the printer what to do. Before a print, a simulation of what the print should look like is run. As these vast robotic arms move gracefully to deposit metal, a suite of sensors captures visual, environmental, and even audio data. Relativity’s software then compares the two to improve the printing process.

Along with the use mentioned above, the use of AI for space exploration extends far beyond the garages. Now they are being used by the likes of NASA to explore galaxies and stars. Even the famous mars Curiosity rover is equipped with an AI toolkit, which autonomously tests the rock samples without human intervention. From solving computational fluid dynamics problems to landing the rockets, from computers to launchpads to extraterrestrial missions, AI has found itself in every stage of the rocket building process.

The post HOW IS MACHINE LEARNING HELPING BUILD SPACE ROCKETS appeared first on Artificial Intelligence.

3D printing is already being used to produce electric bikes, chocolate bars, and even human skin. Now, a new AI algorithm that learns each printer’s imprecisions can tweak print jobs to ensure greater accuracy.

The engineers who developed the algorithm find it increases a 3D printer’s accuracy by up to 50 percent. That can make a big difference for high-precision industrial jobs, says Qiang Huang, an associate professor of industrial and systems engineering at the University of Southern California, who helped create it.

Industrial 3D printers and additive manufacturing devices often use expensive materials, so the cost of throwing out prints that aren’t quite right can add up. For instance, it can cost hundreds of dollars in materials to 3D print a single airplane part. And some printers require seven to 10 drafts per job in order to fabricate an object that is accurate in every dimension and curve.

Huang’s group’s program, PrintFixer, requires a given printer to only produce five to 10 draft objects during initial setup. Once those objects are 3D laser scanned and compared to the computer-aided design (CAD) files that generated them, the algorithm develops a neural net model of the printer’s inaccuracies.

Then, when a new print job is sent to the same printer, PrintFixer recognizes the sorts of corners and curves that need to be distorted slightly so that the ultimate printed object is completely true to its design.

Huang says his group actually developed a new kind of machine learning algorithm to power PrintFixer. Existing image processing neural networks were mostly looking at the wrong thing for what his group needed to do.

“For additive manufacturing, you want to know what’s the accuracy, but for image processing you want to know its classification,” he said. “You want to know if it’s a cat or a dog in a different posture. The shape accuracy is not the main concern about the image.”

Instead, they approached the challenge as a sort of real-world integral calculus problem. Each layer a 3D printer lays down is like a narrow portion of the area underneath the curve of some mathematical function. (Say for instance, it’s the 34th successive layer required to print a sprocket.) From looking at the CAD file, they know what the shape of that 34th layer should be. And from the laser scan, they know what that 34th layer actually looks like.

So their algorithm performs a mathematical operation called convolution (not related to convolutional neural networks) that compares this small cross-section of the object with its idealized CAD file representation.

Add up all those convolutions over however many sample objects are scanned in, and the neural net begins to “learn” the quirks of the printer that created those objects. So when a new print job comes in, PrintFixer can compensate for the printer’s quirks and rendering peculiarities.

Huang says his group has already worked with Honeywell Aerospace and Hewlett Packard Enterprise to refine the algorithm. PrintFixer has not, however, been released to the public. “I’m hoping within three years we can make it available, if not sooner,” he says.

Among the complications still to be tackled are the different kinds of inaccuracies that crop up in 3D printing. Some simply pertain to the geometric shape of the object—one portion of the object should be a perfect circle, for instance, but it’s a oblong in one direction—and those are the sort PrintFixer is currently equipped to handle.

However, other inaccuracies come from processes like delamination, in which the layer beneath sticks to the new layer being laid down. And so the gooey sandwich of layers gets distorted in ways that PrintFixer’s neural net is not yet able to resolve.

In the meantime, Huang says he wants to make PrintFixer available to select users who can devote some of their own resources to the project. (His team is not big enough to provide customer support as would be required for a commercial software company.)

Aerospace and dental appliances are two industries that Huang suspects might be most interested in PrintFixer.

“When you 3D print those dental products, you want to learn what you’ve printed and make sure the next print will be very accurate,” Huang says. “So you can reduce the time, you can reduce the number of visits for the patient and also reduce the cost.”

Huang’s group describes its algorithm and the math behind it in a recent issue of the journal IEEE Transactions on Automation Science and Engineering.

Of course, now that they’ve developed this new machine learning algorithm, Huang says there may be other applications for PrintFixer’s novel neural net design beyond the realm of 3D printing. In fact, one or more of his PhD students are interested in this question.

“There are a few possibilities in my mind,” Huang says of new applications for PrintFixer’s machine learning algorithm. He wasn’t, however, able to offer specifics about those applications, because the work is still in a very preliminary stage. “That’s really what the research is for,” he says.

The post 3D Print Jobs Are More Accurate With Machine Learning appeared first on Artificial Intelligence.

In an application filed with the US Patent and Trademark Office, Facebook is exploring the development of a `modular electromechanical device` (read smartphone) which will allow users to add different components onto a device.
The modular device can incorporate a speaker, microphone, touch display, GPS and function as a phone, Business Insider reported late on Friday.
Facebook`s hardware lab `Building 8` which is focused on developing cutting-edge camera and machine learning technology is working on the project.
The patent noted that millions of devices connected to a server could be loaded with different software based on components that are swapped out.
“Typically, the hardware components included in the consumer electronics that are considered outdated are still usable. However, the hardware components can no longer be re-used since consumer electronics are designed as closed systems. From a consumer prospective, the life cycle of conventional consumer electronics is expensive and wasteful,” the patent read.
The device could be made using `3D printing` technology to function as a phone or a music speaker, the report added.
Not just a `modular` device, Facebook is reportedly foraying into consumer hardware products that may involve next-gen cameras, augmented reality (AR) devices, drones and even a brain scanning technology.
At its `Building 8` facility, the company is working on at least four unannounced consumer hardware products.
Tech giants like Google and Apple have also been exploring this area.
However, Google suspended its ambitious modular `Project Ara` last year.
The Ara team developed a concept design that reimagined the smartphone as a series of smaller, LEGO-style bricks that could be attached, rearranged and swapped out in seconds, media reports has said. ade.

The post Facebook coming with ‘modular’ smartphone? appeared first on Artificial Intelligence.