Large-scale automated production faces the challenge that theoretically identical machines, even of the same brand and model, exhibit subtle variations in their operation. These differences, or 'noise,' can accumulate over time and lead to significant manufacturing defects, affecting reproducibility and reliability, especially in high-precision sectors like aerospace or architecture.
To address this issue, a team from IMDEA Materials Institute, in collaboration with Lawrence Berkeley National Laboratory (USA), has created an intelligent algorithm. This system analyzes performance differences between machines to determine the most suitable optimization strategy, whether joint or individualized, based on the detected similarity.
The new method first performs a diagnosis to profile each machine. If they are very similar, it applies a joint optimization to maximize efficiency. If significant differences are detected, it activates an individualized optimization, prioritizing precision and avoiding biases.
The method's validation was conducted with three theoretically identical 3D printers. The algorithm detected appreciable differences and determined that each required a specific optimization strategy, resulting in faster convergence and a substantial reduction in errors in the weight of printed parts compared to a uniform approach.
Researchers highlight that the system learns the operational 'personalities' of each machine and uses them to its advantage, determining whether it is more efficient to treat them as a team or as individuals. This improves precision and saves resources by avoiding failed experiments, a key advancement for the full automation of laboratories and factories.
Although the study focused on 3D printing, the methodology is applicable to other high-performance fields such as materials discovery, chemical synthesis, or sensor calibration. The work was carried out by researchers from IMDEA Materials Institute and Lawrence Berkeley National Laboratory, and received funding from the Community of Madrid, Spain's Recovery, Transformation and Resilience Plan, and the European Union.
