Japanese Startup Makes Smart Masks for 8 Language Translators


TEMPO.CO, JakartaWhen the Covid-19 pandemic made face masks a daily necessity, Japanese startup Donut Robotics saw an opportunity. They created smart masks with high technology designed to make communication and social distance easier.

Linked to an application, the C-Face Smart mask can write utterances, amplify user voices, and translate speech into eight different languages.

The trunk at the front is very important for breathing, so the smart mask does not offer protection against the corona virus. Instead, it was designed to be worn over standard face masks, said Donut Robotics CEO Taisuke Ono, as quoted by CNN, Tuesday, August 4, 2020.

Made of white plastic and silicone, the mask has an embedded microphone that connects to the user’s smartphone via Bluetooth. This system can translate Japanese and Chinese, Korean, Vietnamese, Indonesian, English, Spanish, and French.

Donut Robotics first developed translation software for a robot named Cinnamon – but when the pandemic struck, the robot project was put on hold. That’s when the team engineers came up with the idea to use their software in face masks.

Donut Robotics began its development in a garage in Kitakyushu City, in Fukuoka prefecture, in 2014.

Ono founded the company with engineer Takafumi Okabe with the aim of “changing the world with small and mobile communication robots.”

With venture capital investment, both of them applied to the Haneda Robotics Lab – an initiative that looks for robots to provide services for visitors at Tokyo’s Haneda Airport.

According to a spokesman for the Haneda Robotics Lab, robots fill the need because Japanese labor is declining, making it more difficult to recruit human staff.

Cinnamon Robot Donut Robotics – designed to provide useful information and help tourists navigate the airport – was one of four prototype translation robots selected by the project in 2016.

Haneda Robotics Lab said the Cinnamon robot won the competition because of its attractive aesthetics and userism and friendly design, and because the translation software performed well in noisy environments.

This success prompted the company to move to Tokyo and take on three new team members.

Ono said the Donut Robotics software uses machine learning developed with the help of translation experts and specializes in Japanese.

He claimed that “this technology is better than Google API, or other popular technology” for Japanese language users, because most competing applications focus on translating to and from English.

The team began testing the prototype at Haneda Airport in 2017 and continues to develop the technology.

But earlier this year, Covid-19 hit Asia and the airport project came to a standstill. “We ran out of money and wondered how to keep the company going,” Ono said.

The team looked for solutions and came up with ideas to adapt the software to products that would sell in a pandemic.

The corona virus pandemic has caused a boom in the sale of face masks, by wearing masks in public now mandated in many countries in the world.

Seeing the opportunity to monetize their translation technology, Donut Robotics launched a fundraiser on the Japanese crowdfunding platform Fundinno in June. They collected 28 million yen (Rp. 3.8 billion) in 37 minutes. “It’s very surprising,” Ono said, “because it usually takes three or four months to get that kind of money.”

The second round of crowdfunding at Fundinno in July generated 56.6 million yen (USD 7.8 billion), which Ono will use to develop translation software for the international market. To increase production, Donut Robotics has partnered with a company in Tokyo.

Ono said the first distribution wave was expected to take place in Japan, with 5,000 to 10,000 masks available in December. They will be valued at $ 40 to $ 50, he said, by subscribing to additional applications. Donut Robotics will not expand overseas until April 2021 at the earliest, but there is interest in Britain and the US, where they plan to crowdfund on Kickstarter, Ono said.


Source link


Please enter your comment!
Please enter your name here