Bringing Artificial Intelligence into Additive manufacturing

Even though the additive manufacturing process is all about adding layers to make a product/component, any 3D printing process can only be carried out by making the dependant parameters optimal. Any 3D printed product must confirm to requirements with regards to time for print, quality and mechanical performance. The usual way of resorting to trial and error, a method to find out what the lattice parameters are and the optimal print conditions are is tiresome and doesn’t qualify as a suitable method. This is where machine learning comes into play. It is used for generative designs and pre fabrication testing. The whole aim of machine learning is to improve printing efficiency and lead to cost savings. Additive manufacturing and Artificial intelligence can go hand in hand with an aim to create service oriented production processes used in the industry. Artificial Intelligence is currently used in additive manufacturing for two purposes mainly:

? Improving efficiency in pre-fabrication stage

? Defect detection

According to an article in engineering.com, by Stefan Krauss, Global General Manager, and Discrete Manufacturing Industries at SAP:

“Manufacturers who can take advantage of ML to predict when equipment and parts will fail, then subsequently employ 3D printing to proactively print and ship replacement parts ahead of these failures, will enjoy significantly reduced spare parts costs and delivery times, and higher customer satisfaction.”

The future applications

1) Real time control

Defects can be identified by computer vision and machine learning can be used to dynamically control the printing process to avoid defects. Machine learning based in-process detection process to control quality can help in limiting waste of materials and time. The powders are very costly when it comes to 3D printing. 3D printing of Titanium which has diverse applications has issues regarding near net shaping which can be mitigated with machine learning controlled manufacturing.

2) Predictive maintenance

Usage of machine learning in 3D printing of spare parts industry. According to a 2017 report from Price Waterhouse Cooper, over the next five years, a significant proportion of today’s spare parts providers agree that 3D printing will play a dominant role in the spare parts business. The report also claims this is a potential candidate for usage of machine learning techniques in predictive maintenance and in the spare part 3D printing process. In a manufacturing setup that is discrete machine learning can be used to the life of a part in operation. Machine learning could also be used to proactively identify time for part replacements by using predetermined replacement schedule data.

The biggest applications AI in Additive manufacturing till date seem to be in improving designs and promoting process efficiency. But the future is bright and AI will help AM in following ways:

-Printability checks

- Making a complex printing process simple

-Cybersecurity