An optimal alarm system can robustly predict a level-crossing event that is specified over a fixed prediction horizon. The code contained in this package provides the tools necessary to design an optimal alarm system for a simple stationary linear dynamic system driven by white Gaussian noise.

Given a specific modeling paradigm that can be learned using appropriate data-driven or machine learning techniques, an optimal alarm system can be designed to elicit the fewest false alarms for a fixed detection probability. When the modeling paradigm is a simple stationary linear dynamic system driven by white Gaussian noise, it is easy to use the well-known Kalman filter to enable value prediction of future process values. During implementation, these value predictions can subsequently be used to enable the prediction of associated level-crossing events that may occur in the future, using a parameter that was selected to achieve robustly optimal performance during the alarm system design stage.

This work was done by Rodney A. Martin of Ames Research Center. NASA invites companies to inquire about partnering opportunities. Contact the Ames Technology Partnerships Office at 1-855-627-2249 or This email address is being protected from spambots. You need JavaScript enabled to view it.. Refer to ARC-16561-1.