There are several measurement methods for objective evaluation of speech intelligibility in rooms with sound and speech communication systems. The Articulation Index (AI) is one of the first methods and was developed by Bell Telephone Labs already in the 1940s.
The Articulation Index assumes that intelligibility in speech communication systems is composed of the sum of the intelligibility of individual frequency bands of the speech range. Therefore, in the AI index, the speech frequency range is divided into many frequency bands, octaves or thirds, and the characteristics of each are determined. In addition, the signal-to- noise ratio is also determined and weighted for each frequency band. The sum of the transmission characteristics is the Articulation Index.