See the original story in Japanese.

When I go to startup events both in Japan and overseas these days, I am usually faced with a panel discussion focusing on robotics where Asimov’s Three Laws of Robotics comprise the topic that always comes up. But I think the laws should not be applied to robotics only. It is obvious that various products created by startups have to adopt them as well for the sake of people.

Born out of the third batch of the Docomo Ventures incubation program last year, I see Oton Glass becoming one of the most helpful products globally. Keisuke Shimakage, CEO of Particular Design developing the smart glasses product, is attending the Institute of Advanced Media Arts and Sciences (IAMAS), a public university in the central Japanese city of Gifu, between Tokyo and Osaka. He started developing Oton Glass to support the daily life of his father who was dyslexic.

Dyslexia is a kind of learning disorder where patients have no problem at all with their eyesight, intelligence or ability to comprehend words but suffer from significant difficulties in reading and writing words. In recent years, Hollywood celebrities like Tom Cruise and Steven Spielberg admitted being dyslexic; the symptom occurs in 10% of the entire population in the US and Europe.

Dyslexia inconveniences the patients only when reading and writing words, which leads to a lack of awareness among the general populace. However, it is obviously inconvenient as people use written words for daily communications.

Oton Glass was originally designed to integrate with a character recognition API (application program interface) provided by NTT Docomo, capturing images with a built-in camera, recognize them on the cloud, send synthesized voices back to the user. So the platform can support people with reading disabilities without human assistance, but Shimakage told us that their approach soon reached a technological limit.

We have learned many things through user testing. For example, when waiting for my turn in the hospital, I would want to know if my receipt number is displayed on the board. However, my behavior that tries to point my iPad camera to the target object so that the device can recognize characters in it would be annoying for people around me.

We have heard many voices saying that they don’t want to use it. We found that there was inconvenience that non-handicapped people couldn’t recognize.

When starting development of the product, the team devoted itself to increasing the accuracy of character recognition. However, regardless of how accurate the sensors they adopted for eye-tracking, it was difficult to point out the characters to increase the accuracy for recognition. Furthermore, recognizing characters takes time, so users need to wait for a speech response after gazing at the target object, which led to stressing out.

Under that situation, the team knew of Copenhagen-based BeMyEyes. The Danish startup provides a crowdsourcing platform where the visually impaired can share live video of what they are looking at via a connected camera with helpers online so that helpers will describe it over the voice chat.

While many of BeMyEyes users are based in Europe and the US, the Particular Design team thinks that they can cooperate with BeMyEyes in many aspects including encouraging Japanese-speaking helpers to join the platform as well as enhancing its use case beyond, to helping dyslexic patients.

Shimakage elaborates:

For example, we would like to further develop Oton Glass in collaboration with BeMyEyes’ existing community so that our hardware will be more optimized for their users.

When we recently exhibited our product at a conference focused on accessibility for the disabilities, we heard many voices from glaucoma patients that they want to use it.

So will the Oton Glass platform pivot to a human-powered crowdsourcing from the character recognition-based technology? In response to this my question, the team’s hardware developer Naoya Miyamoto explained:

We will not give up on character recognition technology. Since technologies for eye-tracking and character recognition will further advance from now on, our solution will become a hybrid of both human-powered and machine-based character recognition.

In partnership with the researchers at Osaka Prefecture University who have joined the CREST project with their character recognition technology, the Particular Design team wants to further develop the product, closely working with dyslexic patients, experienced people from academic circles, and other startups.

While many hardware startups typically use Kickstarter or Indiegogo in their early stage as a marketing tool rather than for fundraising purposes, the Particular Design team wants to focus on solving the problems of dyslexic patients participating in user tests before putting funding, production system, and marketing expansion strategies into consideration.

In the Docomo incubation batch, former Mixi CEO Yusuke Asakura mentored the team and gave them a huge impact about where they can head to, so they expressed a huge appreciation to his effort.

We could get a supportive comment from him for the team’s future growth.

I participated in the last year’s batch (of the Docomo Ventures’ incubation program) from the screening process. I thought that Shimakage’s team deserved a nomination because I was so impressed with his passion that he wants to develop a product making his father’s life more comfortable. I think that the product born out of his real experience or motivation will be probably something gentler for people using it rather than if somebody else had done it.

I’m looking forward to the growth of the product so that it will improve the quality of lives of his father and other people suffering from the same syndrome.

Translated by Chieko Frost via Mother First
Edited by Masaru Ikeda and “Tex” Pomeroy