Limited only to 300 essential phrases, it may pave the way for customizing translations for specific purposes
Panasonic unveiled its multilingual megaphone (megalingual?) Megahonyaku last year. Now it’s ready to start selling what might rain really loud translations on unsuspecting pedestrians for the monthly subscription fee of $183, or ¥20,000.
The Megahonyaku, a portmanteau of megaphone and the Japanese word for ‘translate,’ can translate from Japanese into Korean, Mandarin Chinese, and English. Among other things, they want to have the megaphone ready to service the 2020 Olympic Games in Tokyo. The company started testing the service with some 30 organizations back in 2015 according to the English news service Japan Times. They are reportedly targeting 10,000 enterprise customers. Getting there would probably take a while though if they don’t add more languages.
Panasonic isn’t the first company to come out with a piece of hardware that uses machine translation to automate interpretation. Xerox debuted a copy machine feature this year with options in more than 30 languages. But Megakonyaku is far more limited. It only has the capacity for about 300 common phrases for emergency services and others who might use megaphones. Even if it could translate 300 languages, a limited capacity of 300 phrases per isn’t really cutting edge.
It’s not clear what kind of algorithm they are using, so while it is entirely possible Panasonic is using software that can recognize variations of certain phrases and produce a pre-saved version of that same phrase in one of the three other target languages (which indeed would be a feat), it still would not approach the abilities of the open machine translation tools on the web today that allow input of virtually any phrase.
Of course, a limited vocabulary also means the chances of a laughable translation akin to a lazy middle schooler plugging his Spanish homework into Google Translate (¿Estoy la biblioteca, sí?).
Limited abilities, abundant competition
There are a few applications that will translate text found within images via your camera. Waverly Labs made, well . . . waves earlier this year with its crowdfunding campaign for an in-ear device that simultaneously translates languages over the course of a conversation. Another startup, Mymanu, is working on something similar. Microsoft Translator debuted simultaneous group translation in December that will produce a text interpretation of a face-to-face conversation. Skype debuted something comparable the day before.
Metropolises with a high number of foreign residents and visitors would probably benefit tremendously from this sort of device though. Imagine if New York City had been equipped with these back in 2001.
While there is a market for this, it is not clear why Panasonic is electing to use a subscription service business model instead of simply selling sets of Megahonyakus together. Their model and limitations could open the door easily for other competitors to move in using basic machine translation and natural language processing algorithms that cater to a number of languages.
They seem to be aware of this, however. Panasonic also introduced a smaller device earlier this year akin to an electronic name tag that tourists can wear and set up to display translations to conversation partners. At the time, they claimed it also covered Thai and that they planned to add several more languages in the near future. That device though is more versatile in its language options.
It will be interesting to hear how well sales go over the next few months, but expect more companies to try to capitalize on the idea of marrying language software to specific devices beyond megaphones throughout 2017 and 2018.