This podcast is brought to you by BarrierBreak Solutions Private Limited and Score Foundation.
Hi, my name is George Abraham and welcome to this edition of Eyeway Conversations. My guest today is a professor from IIT Delhi. Welcome Professor Balakrishnan, to the show.
Thanks George for this opportunity to talk to your audience.
Let's begin by asking you how did you get interested in Science and what was your journey like?
I grew up in a place called Pilani. And there I actually lived in a campus. My father worked in an institution, at CSIR lab called C entral Electronics Engineering Research Institute. All my father friends and colleagues are all scientists, they were all building electronic systems. So this was my journey from age zero to 16 when I entered college. So it was very much like I enjoyed looking at what they were doing, they used to have annually an open house where they will sort of throw open and all of us will go. Of course, not when we were very young, but when we have definitely turned 7-8 years. We will see what are the type of things they are making in the lab. Finally, I ended up studying Electronics at BITS Pilani. And after I graduated, I had a chance to join IIT Delhi as a scientist. And that was a very important part of my journey, which actually finally took me to assistive technology. Though this lab that I worked on, was not an Assistive Technology Lab. So primarily this lab was working on defense, mostly on naval defense equipment. They were trying to indigenize many of the equipment in defense, which they were actually importing at that time, from Russia, UK, and so on, so forth. But I started my career not just by building prototypes, but actually building systems and which we had to install and show that it actually works on the ships very difficult environment, both working in ships and testing them on ships. But this was part of my PhD journey. And that gave me a very different outlook in life. So I always like to do things which finally start working in the field. How did I get into assistive technology was a very, very chance meeting with Dipendra Manocha. Dipendra met me and at that time, he wanted a screen reader for Emacs. Emacs screen readers at the time were not available. And initially, my journey was very sort of not really direct involvement, I just used to get some project students and introduce them to Dipendra and Dipendra would actually give them some projects. And my job was to just to broadly supervise them, give them the grade so that you know they are happy, and they work on the projects. But the smartcane change all of it. So these smartcane, on board and dot book, these three, in principle, have the potential of being revolutionary.
So can you tell us a little bit about each of these products?
Yeah, so the smartcane journey started, I think, almost in 2005. So after working for a couple of years with Dipendra, he came up with this need for a device which actually can detect obstacles which are knee above and doesn't have a footprint on the ground. So it can actually prevent upper body injuries when people walk with a cane. And they say this was a big challenge. And there are no solution available and in environments that we live in a lot of things which are overhanging this tree branch or an open window or air conditioner room cooler sticking out on a corridor, and these are not visible, so to say from the cane. And so we need a device which is knee above and we started working on it. And of course it's very important part of the journey is this student called Rohan Paul. Okay, so no journey is complete till you have such bright students who are part of your team. So he came with a very interesting background, both parents being invoved in medical science, grew up with a very strong background of giving back to society. There was a team, of course there were people, but clearly he provided the leadership and made in the three years that he started, you know, they started when they were in third year, and by fifth year we had many prototypes, three or four prototypes. And in this journey, we also met Professor PVM Rao, who is the other very strong pillar of the lab. So because we wanted to prototype and 3D printing had just come to the campus and it was being handled by Professor PVM Rao's lab. So this was 2006-7, that time period, and we went through a Professor PVM Rao and he was so happy to sort of help us and give his own ideas of 3D prototyping, you know, became, let's say lifelong partner for us to technology with us. And then we came up with prototypes, but after the prototypes, this part came where we struggled a lot to actually get into production or anything closer to production, because we realized that what students can do we have achieved. Now we had to actually have an industry partner and we advertised and very interestingly, Phoenix medical systems, Shashi Kumar from Phoenix, he actually applied to become a partner. Two people applied, we chose him. And once he sort of saw the product, and we started discussing, he realized that it requires fair bit of investment. We went from one ministry to another at least three ministries, 10 proposals in partnership with Phoenix. Everywhere the challenge was that they would find us but not the industry partner, and then the Wellcome Trust came in.
What about the product onboard? That sounds very interesting too.
So the onboard was funded by tide program, two phases of onboard. So this essentially has visually impaired people to board public buses and we have done extensive trials. First started in DMTS buses in Delhi, but then we did a very major trial in Mumbai. 25 buses of BEST, we installed 21 users and XRCVC was a major partnern there. Sam Taraporewala and his team helped us to get the blind users and so on. Very successful, BEST was very happy with it, they gave us a letter that we will put it on all our 8000 buses. But we're still looking for funding, we couldn't go beyond it. We came back, again did a set of trials in Delhi, miniaturized it. So it's ready to go for production and the only thing is, it's stuck because nobody wants to fund large scale installation. And our idea is unless we do trials with large scale installation, we cannot go for regulation. Finally, it will be a regulatory requirement that the buses should be accessible. That's the only way the buses will invest. Though the investment is not much, 5000 rupees per bus is not a big investment and a user's device will cost 300 to 400 rupees. And we did not realize it but though we developed it for visually impaired, it actually helps a range of users, it actually helps any user who needs help for boarding. This particular device by an audio feedback alerts the driver and the conductor saying that there is a person who needs help, wants to board the bus. And this was pointed out by not by us by the BEST people, Mumbai that you know your device is more universally applicable than what you claim.
So what I wanted to check is that this onboard, is this a one piece device or there's something that the beneficiary or the blind person also has?
We have two solutions. So now we have a universal solution. So we have one is of course the device goes on to the bus, which is a basically a controller and a speaker and an RF communication device. It's installed at a prominent place in the bus, just next to the entrance, front door entry door so that people get an audio cue for boarding. The other device visually impaired person can use a mobile, the challenge is the range will be very small. Yeah, they can only check the bus around 10 to 15 feet when it is but if you want a larger distance, then we have a handheld device which works on RF communication. And is a very small device. So it is you know, just like maybe a slight change.
So there are several buses coming, stopping and going...
This device is a very simple user friendly device. It has only two buttons on it. Yeah, both Braille mark one is a query button another is a select button. So yeah, he or she presses the query button. And the query, what it does is it gets an RF signal, signal to all the buses in the neighborhood hear that and respond with their own route numbers. This device collects all the route numbers, as is all RF and reads out the route numbers one by one on the hat, the device that the user has on the mobile device. Right now the user if the any of the route numbers of interest, let's say it reads out 501 and that's of interest, then at that time, he or she presses the select button. As soon as you press the select button, that particular bus starts speaking out 501 on his speakers, right, that becomes the audio cue for the even if the bus is moving, he can wait further because the people can sense it's moving audio cue is moving. Yes. And each time he or she presses the button, then it will again speak out 501 so he can get oriented towards the entry door and towards the bus. It is a globally unique device. There is no such device, we have a patent on it. So you know everything is nice about it. Very inexpensive.
This is a very fascinating product. So let's move on to the next one, which is the Dotbook. Tell us a little bit about the Dotbook.
Dotbook is a refreshable Braille display. Yeah, so it is it's commercial name is Dotbook. So it's a refreshable Braille display, so we had two versions launch which is 20 cell and 40 cell. 40 cell is a more interesting one because 20 cell there are now market competition is available alternate devices are available at 16 cell and 20 cell with the Braille keyboard. But this 40 cell is with the QWERTY keyboard. So people who are familiar with laptops and so on so forth, can very quickly change over to this and is fully featured. It has email, it has a document editor, it can browse the web all of it, you can do it on them, using the Braille for T cell display. The device is functional, the device is also in low volume production, we did have a problem even after the launch. We had some problem in terms of reliability of past production of sales. Yeah, but let's all now sorted out and you know, first. So we have two orders now, we are servicing one order, we are almost done 25 units, another 25 units is an export order. So that's also going through. So the issue essentially is now we have sort of broken the challenges and you know, it will scale up now we need basically some type of large organization, which will do a sort of big volume model, because producing in small volumes and selling is always a challenging space because costs become very high when you do very small batch sizes. So you need to do large batch sizes, and we are talking to people and you know, it gets into ADIB scheme or some other schemes which are useful. And that 70,000 75,000 rupee that's roughly the price of this device, which is anything comparable is two and a half to three lakh rupees.
So, as I understand it this product, though everybody is now talking about computers, and E text and all that this Dotbook actually enables people to continue to read in Braille...
No. So the point is, I'm not sure how, at any point of time only through audio, a person can learn mathematics, right? No, I don't think it will happen. Yes, audio will replace even for not even for sighted audio is becoming much more common for but if you want to do some serious studies, you still have to write and if you had to write you have to read. Correct. How do you read and write without Braille? So as far as the blind people is concerned, Braille will remain, it's use will come down, it has already come down. But if it gets completely eliminated, I think then a lot of professions and a lot of education will go away for access will go away, because it is still not possible to do a lot of things in purely in audio.
I was reading about the Raised Lines Foundation. So tell us a little bit about this. And what does it actually do?
Raised Lines Foundation is actually a nonprofit startup that we have started, which actually is into tactile books and production of tactile books and tactile diagrams. So the story is quite interesting when we did the smart cane. So at one point in time, we wanted to do manuals, self learning manuals for the visually impaired for smartcane. Yeah. And so our students want to explain how the actually ultrasonic ranging works, as part of the manual, because there are the fundamental thing and they wanted to do diagrams, and then this, we all of us thought that it must be easy. They went from place to place and we realized that how primitive way the diagrams was made. And in contrast, we also had access to because of this Wellcome Trust funded it. And few times, we had to go to London to defend the project to make presentation. So we were visiting Royal National Institute of Blind, RNIB in London very frequently. What we saw there was fantastic tactile, very good quality tactile books in geography. And then we realized that the reason there's absolutely nothing here, so there are blind schools, which have tactile diagrams, but this is all the effort of the teachers there who make this diagrams and whenever they know, they have the time and energy, they make it available to the students, but there's no book available. And actually the most of the blind schools will tell their students that diagrams are outside the scope. You don't have to bother, just bother with the text. Right, which again, you know, science education is not possible. So then we took up a project, MIT was very helpful, they gave us a center of excellence in tactile graphics. And we reduced the cost of producing these diagrams. So at that time, the diagrams abroad were being produced at around 150 rupees per page, we brought it down to 20 rupees per page. We were the first to use 3D printing for making the tactile, you know, masks you can say so the moulds which actually are used for thermoforming. And we created that software and everything associated with it through the center of excellence, worked closely with NCERT, converted some of their science books, science and maths books into this and went to some schools and tried with some of the children. They were very happy. A lot of feedback, a lot of improvements and today, now we have Today somebody wants to actually get a book converted into tactile, they just have to approach Raised Lines Foundation, they will provide end to end support.
It's really encouraging that IIT Delhi and other professional technical institutions are getting into using technology to make life easier for people who are blind and visually impaired. So what are some of the other projects?
We are into three very different and very interesting projects. So, one is what we call mobility assistance for visually impaired. So this is a project which is video based, and uses all latest AI techniques to look at mobility issues in the Indian context. Even when I walk on a, you know, in the campus, which is relatively I would say much more safer and not crowded, and so on so forth. It's not unusual to have a dog sleeping on the ground. So now, it needs to be detected, it's a safety issue because you don't want to go and hit the dog. So, dogs, cows, signboards, and faces, which are familiar, so, we have a device which will take a video stream now, prototypes are there. So there's still a lot of work that has to be done to optimize them to come to a level where we can use it so and it what it will essentially do is your mobile user interface, but the device has a camera and it has a very strong processing unit and it's neural network based so it's a AI based solution. So, they require a lot of competition. But they have been now well trained and we get to know 90-95% accuracy and most of the... In India, the signboards are very often multilingual. For example, Delhi most of the sign books will be Hindi and English. So you want to separate, so most of this OCR will not work because OCRs won't know what language first of all it is and convert some garbage. So we have to separate it into Hindi and English, convert it into and that creates an upgrade the information that is actually required for the visually impaired to move. So this is one project which is prototypes are there. It will of course take at least two years to reach a stage where we can say that users can use it but first set of user trials will start in under three months. Second project is a project called RAVI, Reading Assistance for Visually Impaired. Again, our focus has always been looking at STEM, text issue is solved. So if you have a PDF text document, you can, you know, clearly use the present screen readers and there's absolutely no challenge. But when it comes to equations, when it comes to tables when it comes to diagrams, making it more comprehensible, making it you know, work well is a challenge. And in many cases, the very legacy software has been used. So it's a completely is more latech created document challenges are less. Word has more challenges, if you take a Word doc and convert to PDF, it has more challenges, but if it's a legacy one, some publishers, old publishers' documents, those just don't work. So this we have a set of rules now, which is actually be able to put together in an accessible ePub format. So a lot of work we have done so for example, removing, if on the PDF document, you have a watermark, the watermark has no use from visually impaired perspective, that will gobble it up. How do I identify and remove one? They are running footers or running headers, let's say every page has the same chapter name at the top, that again will gobble up the document because they will come at some point in time. And if the user doesn't know it's actually a header or a footer, it will just make nonsense out of it. So cleaning up all of these documents and making mathematics accessible. So this is one set of projects we are going on. And this is we have again, no first set of not user trials, but at least our own validation trials in a month or two, we will begin. A lot of tools are ready and it will be a web accessible system, you can go to the web deposited PDF and get converted. The third project which is I think is a very interesting project is to make indoor spaces accessible. Now with Google and so on so forth, now, outdoor spaces, mobility has improved for everybody, not just for visually impaired but for sighted also. And these are accessible but what will happen as soon as you go to a public building? How do I reach office? Today everybody is dependent. So what it works on is it works on basically Bluetooth beacons installed at strategic places within the building, an indoor map of the building and the map gets downloaded onto your mobile. And now we also interfaced it to things like Open Street maps and later on to Google Maps. So that means it will be a seamless integration between coming from outside, entering the building and traveling within the building and going to your own floor and own office that you are interested. So this project is at present on pilot trials at RP center in AIIMS and NAB in Delhi, so, we have had the installation and one of the colleges in Delhi University has approached us apart from our own SIT building where we operate. So, this is a very preliminary trials. So and you know, I would at some point come and seek the support of the community because just like either we will do it within RLF if we can improve increase the object or at some point in time, we would actually like to scale it up and create a company out of this effort. So, that actually buildings become accessible because I strongly believe that what happens in the lab will remain in the lab, students will graduate and the project will be wound up, I also have not much time now to retire. And that still will happen. So it has to become, go out of the lab into a company or you know, for-profit or nonprofit that's of course the challenge.
Thank you very much for this time that you've given us Professor and more power to you and I hope more and more people like you come forward so that young people in our technology space take up you know technology for people with disability as their vocation in life. It will be wonderful. So thank you very much for your time.
This podcast was brought to you by BarrierBreak Solutions Private Limited and Score Foundation.
