NASA, speech recognition and Commodore

NASASpeech_thumbSome time ago I wrote about how NASA used the Commodore Amiga and kept on using it well into the first decade of this century.
The Amiga however was not the only Commodore computer NASA was interested in it seems, as back in 1990 they sponsored research on voice reaction times and speech recognition on… a Commodore 64.

When it comes to voice/speech recognition, it makes perfect sense for NASA to be heavily involved.  Astronauts on spacewalks or in the International Space Station (ISS) work in noisy environments doing jobs that often don’t leave their hands free to control computer systems. Voice-recognition programs, that can block out background noise, help in triggering tasks the astronaut wants to perform. Research on this topic goes back a long time, and it seems the Commodore 64 was still a computer that had not passed its shelf-life when in 1990 Washburn and Putney of Georgia State University published their NASA-sponsored paper called “Voice reaction times with recognition for Commodore computers”.

The authors themselves put it as “The Commodore 64 (C-64) and 128 have been popular computers with numerous strengths (…) as laboratory instruments. With the steady decline in cost of more powerful computer systems, many of the advantages offered by the inexpensive Commodore systems may seem to have diminished. However, the availability and relatively low cost of these computers and of their associated software and peripherals remain tenable reasons for selecting Commodore systems as dedicated data-collection and control units.”

In a series of experiments, using a Commodore 64 and a modified Voice Master cartridge, they demonstrated that the breadbox along with the Voice Master software performed remarkably well in recognizing the test subjects answers after being “trained” to listen to the test person’s voice.
In quite an interesting further experiment they demonstrated that the Commodore 64 could even be used in environments with some background noise present (it involved two monkeys doing tests on a computer and the system had to distinguish the monkey’s “talking” from the noises and bleeps coming out of the computer program they were “playing” on).


The results of all the experiments were so surprisingly good, that the authors even went as far as to state “Although it is premature to interpret the meaning of the observed patterns of vocalizations, the implications for the reliability of the hardware are unequivocal.”

So, did NASA pursue on the Commodore route and put the beige beast in the ISS? No, this didn’t happen and Commodore would no longer be around a couple of years later either, but the research did show that the old machine along with its ease of use, flexibility and wide range of cheap peripherals was capable of doing some remarkable stuff, 8 years after its release on the market.


NASA did continue (and still does) its research on the most optimum voice and speech recognition systems and has ventured into the realm of science fiction (although it’s becoming a reality) with “subvocal” recognition, in which a set of electrodes are attached to the skin of the throat and, without opening the mouth or uttering a sound, the words are recognized by a computer.

Share This Post


Leave a Reply

Your email address will not be published. Required fields are marked *