Hello Annie

Anne Sullivan — the unschooled girl, nearly blinded by trachoma — became one of the most brilliant educators of America and the lifelong companion of Helen Keller. She was fondly called Annie. Founders of Bengaluru-based start-up Thinkerbell Labs didn’t have to think twice before naming their Braille educator device after her.

“In the blind’s world, when you mention Anne Sullivan, they immediately imagine a teacher,” says Dilip Ramesh, co-founder and CTO of Thinkerbell Labs. A few years back, Annie was only a weekend hack for four engineering students — Ramesh, Sanskriti Dawle, Aman Srivastava, and Saif Shaikh — of Birla Institute of Technology, Pilani (Goa campus). But today, the device helps 24 students of Rajyakrit Netrahin Madhya Vidyalaya, a school for the visually impaired children in Ranchi.

In a regular school, when a teacher says, “A for apple”, the lesson is conveyed immediately to the students. However, for a visually challenged student, the tutor has to hold her hand, move her fingertips over raised dots on her Braille board and tell her that the pattern forms ‘A’, to introduce her to the first alphabet.

Often, teachers have little time to spare for such a patient exercise. Even in a class strength of five, each student gets just eight minutes of the teacher’s time during a 40-minute session. While the number of qualified Braille teachers itself is limited, teaching efficiency is also hampered by the tools available. Most special schools in India use primitive instruments such as the Perkins Brailler (typewriters used to write in Braille), wooden blocks (they have Braille alphabets carved into them) or Braille slates (used by the blind to write text they can read without help).

India’s first self-teaching Braille device, Annie, is miles ahead. It is hardware built from Raspberry Pi (a low-cost, small-sized computer) – with a digital Braille keyboard, Braille display, and a speaker – to help a visually impaired child learn to read and write. It comes with audio-guided lessons and interactive content.

Teachers and students are given Annies, all connected via a local server. When the teacher sends instructions via the keyboard on her device, students get an audio notification and can read the inputs on their Annies using its haptic pads.

This replicates a regular classroom where information is disseminated to all simultaneously. Annie’s companion app enables a teacher to customise or create content, while an analytics platform, Helios, helps teachers monitor a child’s progress. At present, the start-up has 35 working devices in India (20 in Ranchi and 15 in Bengaluru) and another 20 in the UK.  

When the team was starting out with this, as a pet project (Project Mudra, then) in 2014, they took the device to schools in Goa. Despite their primitive model – with cells popping out to an alphabet song – the kindergarten students loved it. It was an eye-opener.

“These days a two-year-old opens YouTube and plays whatever he wants. So, they are always taking in information even if there is nobody to give that information to them. That is missing for the visually impaired,” says Ramesh. To dive deeper, the team took Braille lessons in 2015 at the LV Prasad Eye Institute, Hyderabad.

The same year, they did more trial runs in schools in Hyderabad and Goa, and Dawle says the biggest learning was from taking regular feedback from its children users. “A simple feature such as giving the student stars for completing an exercise – a feature we added in just two hours — made children stick,” she says. It introduced an element of competition and collaborative learning, which happens so often in regular schools. “But Braille learning had so far been an individual learning process,” she explains. The team also realised that the solution must include the needs of sighted stakeholders such as teachers, school administrators, parents, and government officials too.

After six to seven major iterations over two years, Thinkerbell included various new features such as two haptic pads — one pad has bigger Braille symbols to help children recognise the alphabet; the second pad helps students train in the standard-size Braille.

After graduating in 2016, the Thinkerbell team moved to Bengaluru, incorporated their company and ran pilots in nine schools across Bengaluru and Dehradun. Meanwhile, the team won LetsVenture’s (an online funding platform) Tech Rocketship Awards 2016, conducted by the UK government’s Department for International Trade. This honour introduced them to the UK market. They even got a chance to show Annie to Prince William and Catherine, Duke and Duchess of Cambridge!

Braille literacy levels in the UK are abysmally low at a mere 4%. The Thinkerbell team thought that Annie may, therefore, have an international market. “Dehradun has the biggest university for the blind in India (National Institute for the Empowerment of Persons with Visual Disabilities). Most blind schools in India follow the curriculum it prescribes. UK has a similar institute called the Royal National Institute of Blind People (RNIB). Its curriculum, called hands-on Braille, is adopted by teachers in the UK and other Commonwealth countries,” says Ramesh. The Thinkerbell team took the hands-on Braille guidelines, adapted it to Annie, and ran pilots in the UK schools. While the company has already sold a few units there, many others have expressed an interest in buying Annies too, according to the start-up team.

In July 2018, their first batch of 20 devices was deployed in Rajyakrit Netrahin Madhya Vidyalaya in Ranchi for grades one to three with the help of Rai Mahimapat Ray, district magistrate, collector and deputy commissioner of Ranchi. “Annie is a meaningful change in the lives of those who need it most,” says Ray. Annie teaches English and Hindi Braille through Hindi audio.

“We are working with LV Prasad Eye Institute and Braille experts from RNIB to build English content solutions for the Indian scenario. We have built our own content (the gamified interface with the star-rating based feedback for instance),” says Ramesh. “It is easy to localise content. Say, the team wants to deploy this in Karnataka. We only need to change the audio and encoding on the software side, and we can teach English or Kannada Braille,” he adds. Dawle adds, “For six-dot Braille, you can have all languages while just changing the software and audio.” Localisation is possible for not just languages, but dialects too. 

“As of now, we see children coming in for extra hours after school to sit with the device,” says Dawle, who adds that, after Jharkhand, governments of UP and Telangana have reached out to them too.

The start-up raised an angel round of Rs.13 million in July 2017 from the Indian Angel Network (IAN) and Anand Mahindra. It is short of $300,000 (approximately Rs.21 million) to closing its next round targeting $1.1 million (approximately Rs.77 million) to speed up sales and manufacturing. Its investors in the latest round include IAN and LetsVenture.

“What appealed to me was its unique intellectual property. Equally important is that it has a global market,” says Rajesh Navaneetham, IAN. “It is the first of its kind in the world (as a device that teaches how to read and write in Braille). Such a solution did not exist even in developed countries such as the US or the UK, and it immediately appeals to all of them. The social impact of it was icing on the cake,” he adds.

According to the Thinkerbell team, Annie is superior to its immediate competition Taptilo, a Braille tutor device by Ohfa, a Korean ed-tech start-up. The latter only teaches Braille Grade One, while Annie also covers Grade Two (grading as per the American Council of the Blind). Further, Annie also teaches students how to type, while Taptilo doesn’t. 

In the next phase of their growth, they aim to have a commercial presence in the US and the UK by the end of FY20. “We have had advanced-level discussions with RNIB for distribution in the UK and USA-based LoganTech for distribution there,” says Srivastava, co-founder at Thinkerbell.

The start-up expects to see a combined revenue of Rs.25 million from India and foreign markets by end of FY20. The team refuses to divulge current revenue details. The device costs around Rs.40,000-50,000 a piece but they have two revenue models for different countries. Thinkerbell charges $940 upfront for the device and $150 as a yearly content subscription fee in developed countries (presently only in the UK). Annie retails for $625 in developing countries with zero subscription fee. Thinkerbell has got its second order from Bengaluru-based social enterprise, SELCO Foundation for a school in the city, where it will deploy 15 devices. 

In the long run, Thinkerbell also plans to come up with solutions for the assessment of Braille skills. The idea is to offer a standardised system of learning for all Annie users, in terms of content quality and learning experience, irrespective of language or location. It could then issue certifications that could help with employment, says Dawle. Navaneetham too believes the solution can go beyond teaching and learning Braille. “There are a lot of products such as note-takers for the visually impaired. Is it possible to expand Annie’s utility to more similar areas? Then that would be a bigger opportunity,” he says. 

The device presently has internet connectivity. Since it allows for reading and writing, Navaneetham is betting big on Annie’s ability to not only help learn Braille, but also generate and propagate more content for the visually impaired.

Literacy is not the final goal for Annie. The Thinkerbell team hopes to enhance the employability of visually impaired people worldwide and even target other disabilities at a later stage. “For a significant section of people, Braille is not just about education, but also about independent interaction with the world. We look forward to being there for them,” Dawle says.