AI detects individual 'personalities' of 3D printers to reduce errors

Even machines of the same make and model can exhibit subtle performance differences, leading to accumulated manufacturing defects in large-scale automated systems like 3D printing farms. To combat this, researchers from IMDEA Materials Institute, in collaboration with Lawrence Berkeley National Laboratory, have created an intelligent algorithm designed to detect and account for these individual machine "personalities."

The developed system, detailed in Advanced Engineering Informatics, analyzes these subtle operational variations to tailor optimization strategies for each machine. This approach aims to reduce errors and enhance the quality of products manufactured in parallel production environments.

The algorithm first creates a unique performance profile for each machine through statistical analysis, quantifying variability. If machines are found to be sufficiently similar, a joint optimization strategy is applied for maximum efficiency. However, if significant differences are detected, the system activates an individual optimization strategy for each printer, prioritizing accuracy.

In a validation study using three ostensibly identical 3D printers, the algorithm identified measurable differences, correctly determining that each printer required its own optimization strategy. The results demonstrated significantly faster convergence and a substantial reduction in errors compared to treating all machines as identical, highlighting the system's ability to account for individual biases.