Using Machine Learning to Improve Manufacturing Processes


When manufacturing complex products, engineering teams typically need to establish a basic monitoring process for the equipment to ensure stability and productivity throughout the operation. In general, simulating a construction takes longer than the construction process itself, making process monitoring a priority for making design decisions. However, as the complexity of products increases and the necessary tools require more intensive monitoring, effective quality management can become a barrier for manufacturers.

To meet this challenge, Renishaw and Altaïr implemented an Al-based quality assurance process using machine learning and advertisingvAdvanced digital gauging to accelerate product development and production.


During the additive manufacturing process, the laser fuses layers of powder while spectral emissions from the weld puddle are monitored to ensure that each subsequent pass matches the original.

Theoretically, teams could take samples of a construction (or print) and compare them to subsequent prints and determine whether or not a process variation is occurring. The problem with this is that there is an incredible amount of data that is produced when printing. For example, a typical print samples at least 100,000 times per second, or about 20,000 samples during the time it takes to flash.

Usually, this data is filtered and relatively basic statistical models are applied to try to make sense of it. This provides useful information, but due to the sheer volume of data created, it becomes virtually impossible for a human to effectively glean deep insights from the data.

Renishaw took advantage of Altair’s artificial intelligence technology that could flag anomalous constructions and regions by analyzing spectral data in real time, enabling faster part development and more stable production. As machine learning algorithms have become more precise with increased computing power, engineers can now rely on ML technology to make decisions from data without compromising ground efficiency..


The Altair signalAI tool has been used to successfully detect anomalous constructions during the metal additive manufacturing process. Without any prior knowledge of previous anomalous construction models, the software was able to determine the anomalous volume and region, as well as reveal the specific 3D region of interest with further analysis. The machine learning capabilities of the technology actively monitored print run data, so manufacturing teams no longer had to physically compare datasets to find anomalies, resulting in significant cost savings. time and money.

SignalAI anomaly detection software can be used in real time locally or remotely in the cloud and includes preprocessing capabilities, enabling complete model construction and analysis within a single platform .

Implementing data-driven, bottom-up decision making enables design and manufacturing teams to holistically analyze processes to improve an organization’s smart factory practices. Looking to the future, an Al-based manufacturing approach offers endless possibilities for machine improvement, such as identifying the best sensor type and location for an application, and shared learning to improve processes within. of a business with a hybrid cloud approach.

For more information visit:


Comments are closed.