RayTron Co., Ltd.

[Company strength]

Our company develops and creates component technologies in the audio, image, and wireless fields, and we combine them with semiconductor technologies to materialize commercial products. That is our way of creating new markets, meeting customer needs, and supplying products as a fabless company. We also work with Hokkaido University graduate school in the audio and wireless fields, and Osaka University graduate school in the image field, in the form of industry-university collaboration. This is done to promote the development of next-generation products that help people save time, enhance convenience, and create new value.

[Business description]

Our company develops application specific semiconductors (ASIC), designs and develops system LSI, and designs, manufactures and sells electronic devices. Our theme is “audio, video, and wireless.” We have R&D projects on algorithms under industry-university collaboration in an effort to create one-of-a-kind, original component technologies. These R&D projects are focused on voice recognition and scream recognition in the audio field, and human recognition and sign recognition in the video field. In the wireless field our R&D projects focus on lossless wireless transfer LSI for high-definition images. We strive to create new components, and utilize them to bring about new functionalities so that we can contribute to society by providing business models to the market.

[Industry]

Electronic part, device, and circuit manufacturing

[Strength of products/technologies]

Our company succeeded in developing the Voice Magic high-performance voice recognition engine, which is highly noise resistant, for use in a non-virtual environment. Conventional voice recognition components had only 60 to 70% recognition rates with acoustic models, and they always came with a close-talking microphone and a switch. These factors made it difficult to embed them in commercial products. Our Voice Magic has a recognition rate of over 98% with acoustic models in the phrased voice recognition method (when the SN ratio is 20 dB). It enables allows continuous real-time responses without using any communication lines. Our voice recognition system allows the user to speak a few meters away from the microphone in a normal living noise environment. This is done in combination with its Always Listening function, which eliminates the need to press a switch.

[Company structure for market development/overseas expansion]

We have no fabrication facilities, so we have partner manufacturers to establish a mass-production scheme based on the quantity of a received order. Our Voice Magic voice recognition module can be manufactured in multiple languages, and customized to support English, Chinese (e.g., Mandarin, Shanghai, Cantonese dialects), Italian, French, and other languages. Semiconductor products are generally traded in English in the world market, so we have prepared English-language product leaflets, specifications, and instruction manuals. Our products are developed as hardware (LSI semiconductors) to prevent copying in overseas markets, and reverse engineering is not possible.

[Market share/Ranking]

Our voice recognition LSI has a high recognition rate in noisy environments, and offers a completely hands-free Always Listening function. No other product offers these functionalities.

[Awards and media coverage]

[Media]

Mainichi Broadcasting System, “Ikegami Akira no Donaru Journal” (January 3, 2015); Yomiuri Telecasting, “Kansai Joho Net Ten” (October 24, 2014); Yomiuri Shimbun, “Jidai o Hiraku” (September 28, 2014); Nikkei, “Kaden Sosa no Robot” (March 17, 2014); Futabasha, “Hontoni Sugoi! Nihon no Kagakugijutsu zukan” (February 6, 2014); Nikkei Sangyo Shimbun, article on our compact voice recognition module (October 22, 2013); Yomiuri Telecasting, “Kansai Joho Net Ten” (July 4, 2013); NTS, publication on next-generation human interface development (June 11, 2013); Medic Times, medical device industry paper “Medical V” (April 2013); Tokuma Shoten, “Goods Press” (April 2013); Nikkei BP, “Nikkei Electronics” (18th issue, March 2013); Nihon Kogyo Shuppan, our high-performance voice recognition system appeared in an article in “Jido Ninshiki” (November 2012); CQ Shuppan, “Section 5: Word recognition mechanism in video recognition IC” (October 2012); Nikkan Kogyo Shimbun, “Mirai Techno” (August 29, 2012); Nikkei Sangyo Shimbun, “Chisana Kyojin,” about our high recognition rate that surpasses human auditory ability (August 21, 2012); Nikkan Kogyo Shimbun, about our wireless LAN that runs at 4 times faster speed (July 30, 2012); Nikkei Sangyo Shimbun, “Tech Frontier,” on the spread of voice recognition (July 17, 2012); Nikkei Sangyo Shimbun, “Digital/Hard,” on the utilization of VB voice recognition technology (May 16, 2012); Nikkei Sangyo Shimbun, on wireless transmission from recorders and players (January 27, 2012); Nippon Television Network, “PON!” (November 9, 2011); Nikkei Sangyo Shimbun, “Sougo/Digital,” on the development of our image recognition system that detects 10 million signs (June 27, 2011); Nikkei Sangyo Shimbun, woman’s scream recognition/alarm module (December 2, 2010); TBS Television, “Kakumei × TV” (November 8, 2010); Nikkei Sangyo Shimbun, on household appliances operated by voice (August 5, 2010); and so on.

[Authorization/Certification]

Management Innovation Program approval.

[Joint research and development]

Joint development of next-generation products with Hokkaido University graduate school in the audio and wireless fields, and Osaka University graduate school in the video field. This is done to facilitate development of such products (industry-university collaboration). Japan Science and Technology Agency: Lossless wireless transfer system for high-definition images (A-STEP R&D Stage, Commercialization Type, April 2012 to March 2017). Ministry of Economy, Trade and Industry: Development of a monitoring platform having an image recognition system with embedded camera (subsidized development project, April 2014 to March 2016). This is a program for promoting the development and introduction of robot nursing care devices. Small and Medium Enterprise Agency: Development of embedded software to realize noise-resistant, voice recognition with a high detection rate for next-generation UI phase boxes (SME manufacturer trial development subsidy program, July 2013 to June 2014). Ministry of Economy, Trade and Industry: Development of a multimodal monitoring platform that uses a voice recognition system (subsidized development project, April 2013 to March 2014). Part of a program for promoting the development and introduction of robot nursing care devices. New Energy and Industrial Technology Development Organization: Development of self-reliance support communication robot (program for promoting the development of assistive equipment, December 2012 to March 2013). Semiconductor Technology Academic Research Center: Realization of embedded voice recognition LSI with real-time response, reduced power consumption, and robust voice recognition performance in noisy environments (next-generation semiconductor configuration support program, April 2012 to March 2013). Semiconductor Technology Academic Research Center: Scream recognition through feature extraction of women’s voices (Next-Generation Semiconductor Circuit Configuration Support Program, April 2011 to March 2012). Japan Science and Technology Agency: Design and development of extremely low energy consumption type high-speed wireless communication LSI (A-STEP High Risk Type, December 2009 to November 2011). Kansai Bureau of Economy, Trade and Industry: Development and commercialization of autonomous auditory judgment LSI (commercialization support project under the Bureau’s SME/venture challenge support program, April 2006 to March 2007). [Main Clients / Business History (domestic)] Development of application specific semiconductors (ASIC); design/development of ASIC systems. Main customers: Panasonic, Sharp, Mitsubishi Electric, Sumitomo Electric Industries, Renesas LSI Design, Konica Minolta, and many others.

[Exhibition History/Information]

New Value Creation Exhibition 2015 in Kansai: 10th SME Expo Sougouten, held by the Organization for Small and Medium Enterprises and Regional Innovation, in the 1st/2nd wings of Intex Osaka, from May 27 (Wednesday) to 29 (Friday), 10:00 to 18:00 (17:00 on the last day). Our booth was located in the special exhibit corner.