Rapid and accurate detection of trace quantities of hazardous and related chemicals can greatly enhance safety and save lives in a variety of fields from military theaters, to law enforcement and public safety. A wide range of sophisticated spectroscopy systems are being continually developed and improved to meet this goal, but they all share a similar challenge: how to rapidly detect trace levels of threat chemicals within a spectrum that may be compromised by Nise, background, chemical interferents, or a number of other effects that vary depending on the instrument and application. Each hardware developer may spend time and resources developing an algorithm to solve this problem, and in the worst cases this process can slow the development of the instrument or compromise the measured performance. Therefore, Alakai is proposing the development of the AgNstic Machine Learning Platform for Spectroscopy (AMPS) that can be trained and deployed to rapidly process and accurately identify threat chemicals in spectra obtained from a wide variety of spectroscopic instruments.
