Skip to content

Tackling Additive with Data - How Crunching the Numbers Increases Adoption and Enables Scale

  • today
  • access_time 12:00 - 12:25 PM CT
  • blur_circularConference
  • monetization_onPaid Upgrade

We are proving that by collecting, analyzing, and applying highly contextualized, in-situ data we can overcome the most one of the most prolific problems in the AM adoption. This being the high complexity and cost of adopting AM systems specifically, the lack of deep expertise, manpower, and support that is required for successful AM adoption and optimization.

Our strategy is simple, we can manage what we measure, so we measure what truly matters: the final part and how it compares against the CAD intent. Our main sensor for measuring final parts is an on-axis and distortion-free camera system. From this we can currently capture printed part data at a resolution of 42 microns in the XY plane and we are currently working our Z depth sensing at 10 microns accuracy. We align this data with the print instruction, CAD intent, all other time-series data generated during the process, as well as any other consumable periphery context from feedstock to weather data.

Since AM generates so much data, we believe that most context processing must happen in real time, on the edge. Our starting point is using the image data to build time series digital twins. We use this data to verify parts and processes so companies can trust and have verified parts directly off the printer.

Additionally, we use this data to create what we call Dynamic Machine Code to provide extremely precise commands that both reduce failure rates and enable repeatability across different machines and locations.

The presentation will feature case studies and tests done in-house and with our research/integration partners. Additionally, expanding into how our technology works, how our data is used, and our plans for future development. Specifically, we will go more in-depth about data registration, computer vision, data analytics, machine learning, and application use.

Learning Objectives:

  • Participants will be able to describe, with first principals approach, what is holding AM back from broader proliferation.
  • Participants will be able to communicate a way to perform part/process validation at a reasonable price and speed.
  • Participants will be able to describe how measuring final parts in-situ is critical for building closed-loop machines and process controls with the goal of building trust in AM end-use parts.