The thermal 3D software I created was sufficient to convey to a broad enough audience that immediate change was necessary to resolve a suspected, now proven, phenomena detrimental to the print process.
Furthermore, this tool packages the context of an entire 3D print job into a low memory footprint that can be archived for future analysis routines.
One major challenge in the art of making is gathering enough overlap in language such that all collaborators feel included in decisions. In order to fulfill this critical need I inserted myself into the contentious boundaries governing the distillation and presentation of thermal data. I extensively adhering to the "keep it stupid simple" principles as I applied my mechanical engineering intuition with understanding of software algorithms to significantly reduce computation and technical complexity needed to answer simple questions.
The expansion of core tools I developed now serve as the primary pipeline for thermal process quality management for the fleet of customer printers. In the future the part agnostic thermal metrics returned can serve as the source data for Machine Learning.
As an example of one early use to measure spatial temperature distribution of power delivery to designated part regions, think of the below x and y profile slice density plots as the temperature equivalent of material density in an MRI or x-ray scan.