Abstract
The scope of this paper is to show multivariate relationships created by measuring and validating multiple sensor data streams. Studies show that 70% to 84% of digital transformations fail (Bendor-Samuel, P, 2019). While there are many causes, they are all attributed to one main concern. Companies and organizations often fail at thinking of the necessary digital infrastructure needed to succeed in digital endeavors. It is common knowledge that if you feed digital models bad data, the output will also be poor. Bad output data translates into unreliable digital processes. This paper demonstrates how fully validated signal data from geothermal facility sensors optimize the algorithms dependent on this data. Individual validations of signals are shown and compared to other data streams related to this process. We hypothesize that while some unqualified data may indicate future unwanted events, other unqualified data are intrinsically bad and will fool future algorithms designed to automate or maintain the geothermal process.
