3D printing is expected to be the future of manufacturing. It is capable of directly creating objects from computer-generated designs without any outsourcing parts. But 3D printing has a high degree of error, such as shape distortion. Yet in the near future, that disadvantage will be overcome by machine learning.
Due to different printers and materials, the real size of the end-products can be smaller or bigger than expectation. Therefore, manufacturers often have to try many iterations of a print before they get it right. What happens to the wrong prints? They are discarded, presenting a significant environmental and financial cost to industry. A team of researchers from the USC Viterbi School of Engineering is solving this problem, with a new set of machine learning algorithms and a software tool called PrintFixer, to improve 3D printing accuracy, helping the process more economical and environmentally friendly.
A recently published work in the IEEE Transactions on Automatic Science and Technology described the “cumulative model of 3D printing”.The team, led by Qiang Huang, an associate professor of systems and industrial engineering, chemical engineering, and materials science, aims to develop an AI model that accurately predicts shape deviations for all types of 3D printing and make 3D printing smarter. “What we have demonstrated so far is that in printed examples the accuracy can improve around 50 percent or more,” Huang said.
PrintFixer uses data collected from previous 3D printing jobs to train AI to predict where the distortion will occur, thereby correcting print errors before a mistake is made. The team aims to create a model which can give the accurate results with the minimum amount of 3D print source data. “We can leverage small amounts of data to make predictions for a wide range of objects.”
The team has trained models that ensure consistent accuracy across a wide range of applications and materials – from metals to aerospace production to thermoplastics for commercial use. The researchers are also working with a dental clinic in Australia on 3D printing of dental models. “So just like a when a human learns to play baseball, you’ll learn softball or some other related sport much quicker,” said Decker, who leads the software development effort development in Huang’s group. “In that same way, our AI can learn much faster when it has seen it a few times.”
The team recommends that users can choose the higher quality 3D printers, in parallel with the use of software to be able to produce the best products.
“But if you don’t want to change the printer, we also have incorporated functionality into the software package allowing the user to compensate for the errors and change the object’s shape”, Decker said. “And then, when they print, they should print with the correct size the first time.” The group’s goal is to provide software tools for everyone, from large-scale commercial manufacturers to hobbyists of 3D printing. Users from around the world will also be able to contribute to improving AI software through sharing the finished print data in the database. “Say I’m working with a MakerBot 3-D printer using PLA (a bioplastic used in 3-D Printing), I can put that in the database, and somebody using the same model and material could take my data and learn from it,” Decker said. “Once we get a lot of people around the world using this, all of a sudden, you have a really incredible opportunity to leverage a lot of data, and that could be a really powerful thing.”