Language is probably the most distinguishable ability that mankind possesses. Human/machine systems have long included written language (for example, in alphanumeric displays and keyboards), but spoken language has mainly been used for communication between people. Over the past few decades, automatic speech generation and recognition by machine now offers a possible alternative to other forms of input and output to computers. However, there are a number of limitations that need to be overcome. Knowledge about how to use speech recognition is developing very fast; in some special areas, it can be a very good alternative, particularly in control room situations.
An interactive speech system consists of speech recognition devices for control or information input and speech generation devices as a form of information display.
Automatic speech technology is of great interest already today, but there are many problems that need to be resolved. In 1985, Simpson et al. gave a comprehensive review of system design for speech recognition and generation. From the human factor/ergonomic perspective, a speech recognition system consists of a human speaker, recognition algorithms, and a device that responds appropriately to the recognised speech:
