Researchers from IMDEA Materials Institute have developed an artificial intelligence algorithm designed to improve automated manufacturing by identifying the unique operational characteristics, or “personalities”, of individual machines, even when they are the same make and model.
According to IMDEA Materials Institute, the system was developed in collaboration with Lawrence Berkeley National Laboratory and aims to improve reliability in automated production by analysing subtle differences between theoretically identical machines before selecting the most appropriate optimisation strategy. The research has been published in Advanced Engineering Informatics.
The researchers said one of the key challenges in large-scale automated manufacturing is that identical machines do not always perform in exactly the same way.Â
These variations can accumulate over time, potentially resulting in manufacturing defects and reduced reproducibility, particularly in high-precision industries such as additive manufacturing for architecture and aerospace.
To address this, the AI algorithm first assesses each machine individually to create a performance profile using statistical analysis. It then determines whether the machines are similar enough to be optimised collectively for efficiency or whether each requires an individual optimisation strategy to improve accuracy.
To validate the approach, the research team tested the algorithm using three theoretically identical 3D printers. While the machines were expected to perform similarly, the system detected measurable differences between them and concluded that each printer required its own optimisation strategy.
According to the study, analyses of the printed components showed clear differences in performance between the machines, with the researchers finding that individual optimisation provided the most suitable approach for the case study.
The researchers said the findings demonstrated “significantly faster convergence and a substantial reduction in errors in the weight of printed parts compared with treating all machines equally and thus failing to correct appropriately for individual biases.”
They added: “Even mass-produced machines may have their own operational ‘personality’. Our system learns these differences and uses them to our advantage, determining whether it is more efficient to treat them as a team or as individuals.”
According to the researchers, the approach could improve manufacturing accuracy while reducing wasted resources.
“This not only improves accuracy, but also saves resources by avoiding failed experiments, a key step towards the fully automated laboratories and factories of the future,” they said.
Although the study focused on 3D printing, the researchers said the methodology could also be applied to other high-throughput experimental fields, including new materials discovery, chemical synthesis and sensor calibration.
The study was conducted by Dr. Christina Schenk, Miguel Hernández del Valle, Luis Calero and Dr. Maciej Haranczyk from IMDEA Materials Institute, together with Dr. Marcus Noack from Lawrence Berkeley National Laboratory.Â
The work received funding from the MAD2D-CM project, the Community of Madrid, Spain’s Recovery, Transformation and Resilience Plan, NextGenerationEU, the U.S. Department of Energy’s Office of Science through the CAMERA program, and the Spanish Ministry of Science and Innovation.
