I recently got a chance to attend the Atlantic Design and Manufacturing Expo (ATX East) in New York City. The expo offered a lot of great information about the future direction of 3D printing and robotics in manufacturing. Here’s a few of the highlights below.
3D Printing
During one of the panel discussions, Markforged talked about some of their current and future developments. One of them was Blacksmith. Blacksmith is an AI system paired with machine learn
ing for quality assurance of their production parts. The sensors on their 3D printers capture data and are able to adjust the printer settings in real time to ensure certain quality standards are being met. This technology can also gather information on what changes and features are required for different shaped parts. As this system learns more and more, it can be implemented into a generative design process to produce the highest quality/lowest cost parts. Generative design takes inputs from the user and this can be mechanical requirements, cost, production methods, space, materials, etc. These parameters are then iterated thousands of times to produce truly unique designs. This is the future of design. Here is an example of a design in the photo below. Most of these designs that I have seen feature a lattice like structure that uses less material while still meeting the structural properties. As we aim to be more sustainable these designs will be seen more and more. 3D printing enables these designs to be tested at a quick rate making the design time unaffected.
One of the other topics that a few of the companies discussed was metal 3D printing. This seems to be in the early stages as only a few companies are even starting to use it. The printer manufacturers themselves partnered with a few early adopters to test and develop the printers. One of the buyers mentioned they just now getting comfortable producing parts after 2 years of owning the machine. There is a significant learning curve to even think about purchasing one of these and the resources are scarce. This was expressed in the panel discussion. Most of the resources to learn were on blogs or websites. As this technology advances, I am not sure if a specialized program would be useful or not. Every new material or composite that is printed will require learning the ins and the outs. It may be better to learn the basics about additive manufacturing and a skillset that enables you to learn new technology fast. This is likely true for a lot of the technologies that will become more prevalent in the workplace.
Smart Factories
Smart factories and the internet of things was discussed in a panel and the importance of good data models was stressed. One interesting comment was from a robotics company talking about starting somewhere with building a predictive/preventive model. Part quality inspection was chosen first because a lot of data could be gathered fast. They chose to save maintenance and failure data models for down the road due to the longer mean time to failure. Fanuc’s new Zero Down Time (ZDT) feature may be able to build a maintenance and failure model faster. There machines can be linked to a larger network and the robots will learn from the group’s failures and start to build the model a lot faster. Even if the machine is not within your company, you can still yield the benefits of this system. Maintenance will become more automated and routine with this system and breakdowns can be severely reduced with an accurate model.
