New software platform leverages data from the world’s largest fleet of connected 3D printers

3D printer maker MarkForged is using AI and federated learning within the new Blacksmith software to enable its machines to improve over time.

This, the Boston-based startup claimed, “marks a major step on the journey toward the future of autonomous manufacturing.”

MarkForged X7 industrial printers are used by manufacturers to turn metals and carbon composites into machine components. By using federated learning, each printer can now contribute to a shared prediction model, while keeping all the training data on the device. This helps preserve privacy and protect intellectual property, while still leveraging the data to improve capabilities of all the devices involved.

Collective knowledge

As a new discipline, 3D printing has room for improvement as turning every new design spec into a product is a matter of interpretation for the 3D printer. By using federated learning, a machine can tap into a sort of ‘collective knowledge’ of the fleet – even if it prints a particular part for the first time.

As they are printed, manufacturers can scan their parts simultaneously with an integrated laser micrometer to assess how close the result is to the intended design. Dimensional data is then benchmarked against the intended design files.

“Blacksmith fundamentally changes the way engineers think about additive manufacturing,” said David Benhaim, co-founder and CTO of Markforged.

“For the first time, they will have confidence that their parts will perform as expected without a time consuming process. Blacksmith makes our platform smarter and is the next step on our roadmap to bring the agility of software to the world of manufacturing.”

Blacksmith is available immediately as a software subscription on all cloud-connected Markforged X7 printers.