August 7, 2022
By: David Ordonez
AI and machine learning is commonly regarded as a pathway to the apocalypse that will ultimately plague humankind. As a matter of fact, it’s one of the most useful inventions humankind has created to this day.
Digital manufacturing is one way these algorithmic programs are being used. At the Massachusetts Institute of Technology, researchers have programmed and trained a machine learning model that supervises and adjusts 3D printing errors in real time. The implications of this are astounding. A professional 3D printing operator might make hundreds of models and prototypes before finding the one most ideal for their purposes. Parameters to figure out would include things like printing speed and the amount of actual material the printer should use. All of this is not the most thrifty and ideal way to spend research money. With the use of computer vision, researchers have designed a model that makes this procedure 10 times more efficient. Not only does this make this procedure significantly faster and more efficient, but what has been realized is that the 3D printing job or operation has been replaced by AI. With minimal supervision, this model is able to accurately find the most useful prototype to print and produce.
Intelligence in these machines is what will prove them to be successful. They’re learning a “complex control policy”, allowing real-time evolution in the laboratory as the workplace changes. A neural network ensures all variables of the print are being noticed and optimized, and at the end what researchers have gained is that this method is NOT limited to 3D printers. With the knowledge and experience this model has given, other manufacturing methods and controls can be facilitated with the use of AI at a greater level than it has ever been, eventually even being able to develop objects with unique chemical or electrical factors.
Works Cited
Zewe, Adam. “Using Artificial Intelligence to Control Digital Manufacturing.” Massachusetts Institute of Technology, MIT, 2 Aug. 2022, https://news.mit.edu/2022/artificial-intelligence-3-d-printing-0802.
Editors at IBM. “What Is Computer Vision?” IBM, IBM, 7 Aug. 2022, https://www.ibm.com/topics/computer-vision.
Chen, James. “Neural Network.” Edited by Michael J Boyle, Investopedia, Investopedia, 8 Dec. 2021, https://www.investopedia.com/terms/n/neuralnetwork.asp.
Comments