LINKLET won CES 2022 Innovation Awards
Dec 2021 -- LINKLET is a user-friendly wearable device for livestreaming and video calling. Our patented design and wireless connectivity best serves use for long hours outdoor.
LINKLET won us 3 Innovation Awards for Wearable Technologies, Digital Imaging/Photography, and Streaming.
Official Solution Partner by Ministry of Economy
Dec 2021 -- Japan's Ministry of Economy, Trade, and Industry nominated Fairy Devices as one of the official solution partners.
Fairy Devices will serve the Ministry with industrial digital transformation by our end-to-end solution from device to cloud and AI.
NEXT UNICORNS 2021 by NIKKEI
Dec 2021 -- Fairy Devices ranked in the NIKKEI's NEXT UNICORNS 2021 for wearable device and AI technologies, a widely renowned annual ranking of most promising startups across Japan.
LINKLET is a user-friendly wearable device specialized for livestreaming and video-calling. With both mobile network and Wi-Fi connection enabled, LINKLET connects wearer(s) and audience through Zoom and Microsoft Teams, which provide easier access and familiar interface for users.
Developed by wide variety of patented technologies, LINKLET’s unique design and light weight set wearer’s both hands free, enable use for long hours, and serves audience with more stable 4K videos of very lively first-person view.
That’s why LINKLET is one of the best options for livestreaming of outdoor activities, remote working/assistance even for field workers, tele-education, and many other use cases.
THINKLET is a LINKLET's sister device for industrial use, specialized for many-to-one and one-to-many communication among field workers for immediate remote assistance and sharing and recording of tacit intelligence.
On top of the basic features of LINKLET, additionally, THINKLET is equipped with a strong voice AI, which detects wearer's conversation on spot and effectively cancels noise.
As part of our Connected Worker Solution, THINKLET helps generate truly eligible data for machine learning and enables system integration with existing enterprise database for company-wide digitization.
Connected Worker Solution is our end-to-end enterprise solution for industrial use. The solution covers from edge device THINKLET to cloud and AI and provides remote working/assistance for industrial field workers, helps sharing and storage of tacit knowledge from skilled human experts, and ultimately, utilizes such data for business intelligence and development of AI expert.
Aiming to replace retiring human experts with AI expert and real-time video calling to voice personal assistant (VPA) in the near future, Connected Worker Solution is highly focused on collecting data eligible for AI.
It is possible because of our powerful voice AI and cloud systems as well as device that edge-processes video and voice data to be cleansed and timestamped, sends data at high resolution to the cloud, then converts them into synced image and text data for each timeframe.
Don't go look for a whole troop of vendors for each part of your wanted solution!
Fairy Devices is a global leader of industrial field digitalization solution and one of the rarest providers of one-stop solution from end to end.
Video, audio, and sensor data are captured and processed at our unique wearable device THINKLET while onsite workers get remote assistance from skilled workers. Data are synced and stored at our cloud platform and further utilized and analyzed by machine learning technology for business intelligence and AI expert.
Just one solution, enhance your business with remote working for field workers, data collected for business intelligence, and AI built for your expertise.
Many enterprises complain that their collected data turn out to be not suitable for analysis or machine learning. Hiring a bunch of data engineers doesn't really solve the problem, especially for noisy and unstable industrial sites.
Our voice AI technologies, supported by powerful multi-MICs array on THINKLET device, detect conversation in interest and effectively cancel noises even in the noisy circumstances. Our patented technologies keep data quality of important scenes against the limit from mobile network.
Timestamped real-time, image and text from voice are synced and generate "meaningful" data that AI can recognize and learn.
Data that are eligible for AI -- the first step to great business insights and maximized business potential.
As a global leader of remote assistance solutions, Fairy Devices is helping an industry leader of energy system and power generation industry.
The industry is facing problems of lack of workforce and increasing retirees of skilled workers.
By empowering service engineers with real-time technical assistance from remotely located skilled workers, Fairy Devices realized onsite digital transformation.
Many industry leaders often face a problem of insufficient skilled workers in expanding their business globally. Especially, quality of machinery inspection & maintenance service is hardly guaranteed in many developing countries.
Connected Worker Solution helps workers on site to be connected and remotely trained by skilled workers sitting far in major markets, easily succeeding their knowledge and knowhow to colleagues in new markets.
In the world full of user-created contents, LINKLET's powerful livestreaming functionality provides a useful tool to anyone who wants to share their enjoyable moments.
Whether you are the one in the moment or you are the one who wants to be in the moment, LINKLET's lively first-person view 4K videos and high resolution zoom-in/-out will immerse you in the scene like you are there, together.
Automatic Speech Recognition
Speech-to-text technology powered by deep neural network (DNN). Highly flexible, our ASR technology easily adapts to customized input such as glossaries and sound modeling to recognize jargons and habits of particular industries.
Multi-lingual Translation
NICT's neural machine translation technology for texts from/to 10 different languages*.
*English, Japanese, Spanish, French, Indonesian, Korean, Burmese, Thai, Vietnamese, Chinese
Text-to-Speech
Voice synthesis technology to make texts sound in male & female voice across 10 different languages*.
*English, Japanese, Spanish, French, Indonesian, Korean, Burmese, Thai, Vietnamese, Chinese
Speaker Recognition System
A technology that recognizes speakers of pre-learned voice out of multiple speakers in audio data, regardless of speech.
Attitude and Intentions Recognition
Recognizes paralanguages from speech and classifies attitude of speakers into 4 different classes: positive & neutral, negative/denial, thinking, questioning/doubting.
Emotion Recognition System
Analyzes tone of voice and other paralanguages from speech and to sort speaker's emotion into 5 different categories: pleased, sad, angry, disgusted, and calm.
Extended Front-End Module (XFE)
A library to process online front-end mainly for multi-channel MICs. Including sound location, sound separation, speech extraction, noise cancellation, and echo cancellation are provided.
Multi-MICs Array
A system of multiple microphones composed of 4-32 channels and highly customizable for various types of layout. Combined with XFE library, our multi-MICs array performs a powerful voice recognition in noisy environments.
Noise Detection
Our technologies detect not only human voice, but also machinery voice -- by pre-learning and analyzing machinery noise or background sound, malfunctioning and defect of equipment can be automatically detected before breakdowns.
M. FUJINO
Y. TAKEZAKI
J. KUCHII
J. YAMADA
H. ISHIGURO