Since 1954, there has been a definite need for an automated method of translating communications – as evidenced by IBM’s introduction of its 701 computer that was tasked with crunching Russian text into English.
In the Internet age, the need for language translation is that much more apparent, with free and paid translation services ranging from Babel Fish to Google Translate littering the space.
Still, with the staggering number of non-English websites out there, it’s rather surprising there aren’t more effective language translation services available.
The Pentagon is hoping to improve language translation technology by pumping money into its Boundless Operational Language Translation service.
BOLT is a universal language translation service expected to “enable communication regardless of medium (voice or text), and genre (conversation, chat, or messaging).”
Outlined in DARPA’s 2012 budget request, the U.S. government hopes to put around $15 million at minimum into the program.
In contrast, Google is developing similar technology, with its voice/text translation and pronunciation tools that speak words with an appropriate inflection.
Although the technology seems to be somewhat akin to DARPA’s vision of BOLT, Darpa claims that Google’s version is somewhat amateurish, as the military version is billed as a much more powerful universal translator.
Congress says that the revolutionary product will “also enable sophisticated search of stored language information and analysis of the information by increasing the capability of machines for deep language comprehension.”
To accomplosh this, DARPA plans to teach a machine how to deal with non-programmatic aspects of speech like poor grammar, slang, and accents.
The idea is to “enable machines to carry on multi-modal dialogues with humans and to comprehend concepts and generate responses in multilingual environments.”
The DoD is also working on the Robust Automatic Translation of Speech, which is designed to translate foreign tongues among high levels of noise, like in a public place or crowded bar.
DARPA proposed $21 billion towards the project (RATS) next year, up from $9 million in 2010. Another program the department hopes to put money towards is MADCAT as it moves to get “tightly integrated technology prototypes to military and intelligence operations centers.”
Concurrently, DARPA seems to be slowly moving away from TRANSTAC, a program designed to create an automated translation interface between two operatives that may have “demonstrated a hands-free, eyes-free, two-way translator prototype.”
If translation services eventually do work successfully, the implications for business and travel will be huge. For example, more people will be tempted to bring their mobile phones on vacation or business trips where they can truly come in handy (aside from calling or surfing the Web) while out and about.