Senvol is to demonstrate that its machine learning software can accurately predict the material performance of parts made on a metal wire directed energy deposition machine after receiving funding from the US Navy.
The goal of the project - titled “Additive Manufacturing Sensor Fusion Technologies for Process Monitoring and Control” - is to implement a standardised procedure for assessing additively manufactured parts for quality acceptance and installation using data-driven machine learning algorithms that 'provide the insights needed to achieve target mechanical performance requirements.'
Since July 2025, Senvol has been utilising its Senvol ML software to analyse in-situ monitoring data from various sensor types and various modalities. The project will run through to July 2027, and hopes to help progress the Navy toward achieving qualified, equivalent AM parts from a more flexible and scalable AM supply base. Through this effort, the US Navy hopes to reduce the need for costly and time-consuming qualification and testing, while integrating in-situ monitoring requirements into NAVSEA policy.
During the project, Senvol will use Senvol ML to parameterise the data collected from the in-situ monitoring sensors and compute summary features associated with the specific phenomena that are deemed worthwhile to gather information about. The objective is for Senvol’s machine learning software to accurately predict material performance characteristics from in-situ monitoring data, as well as to choose process parameters likely to produce parts with the desired characteristics.
Senvol President Zach Simkin said: “Quality assurance in additive manufacturing is critical. For a part to be accepted into the supply chain, there needs to be sufficient confidence regarding how the part will perform. Progress in this area continues to evolve, and we believe that developing a consistent approach to analysing in-situ monitoring data – and developing actionable guidance from it – will enable AM users to more readily meet part acceptance thresholds.”
