0:00:03 > 0:00:05Now on BBC News, it's time to Click.
0:00:10 > 0:00:17This week, the latest disability tech with object recognition, sign
0:00:17 > 0:00:29to text translation and snow seat. -- seat. -- the. -- seat.
0:00:44 > 0:00:48Over the past few years, some of the most fascinating technologies we've
0:00:48 > 0:00:52featured on the show have been the ones that help people with
0:00:52 > 0:00:57disabilities. As the world's burst bionic games proved, the
0:00:57 > 0:01:02possibilities now emerging offer so much potential, whether it be in
0:01:02 > 0:01:07disability, sight or hearing, we've seen how technology is tantalisingly
0:01:07 > 0:01:11close but how long before it really starts to impact people's lives for
0:01:11 > 0:01:16real? This weekend sees the international day of people with
0:01:16 > 0:01:20disabilities and that's a great chance for us to devote a whole
0:01:20 > 0:01:25programme to the latest tech developments in the area. We start
0:01:25 > 0:01:29in Rochester in upstate New York, which has the highest number of deaf
0:01:29 > 0:01:34or hard of hearing people per capita in the US. The unique combination of
0:01:34 > 0:01:37this community and the technical Institute in the area is really
0:01:37 > 0:01:45starting to drive innovation. Paul Carter went to investigate.
0:01:45 > 0:01:49Rochester Institute of technology and its constitute, natch, national
0:01:49 > 0:01:53technical Institute for the deaf, are now at the forefront of
0:01:53 > 0:01:55developing and testing new technologies that help deaf people
0:01:55 > 0:02:00to communicate. One technology is UNI, an innovation that helps to
0:02:00 > 0:02:05live translate signed language into text and speech. Alex Davies demo
0:02:05 > 0:02:07the prototype... At Inglis
0:02:16 > 0:02:24I do the signed for check.I want to check my flight.Your flight is on
0:02:24 > 0:02:28time. As you noticed it came up on the screen as I said that. The
0:02:28 > 0:02:32system works by using sensors with two cameras that detect individual
0:02:32 > 0:02:36points on the joints and fingers and renders them into the software to
0:02:36 > 0:02:40interpret the individual signs, a task more, located than it might
0:02:40 > 0:02:43first sound.Some signs are naturally doing it over and over
0:02:43 > 0:02:51again, bag or wear.This bag is a carry on.That's part of the
0:02:51 > 0:02:55challenge and the complexity of our applications because we have to be
0:02:55 > 0:02:59able to filter all of that out, so how does the software know you're
0:02:59 > 0:03:04not signing bag, bag, bag, just bag. It's safe to say assistive
0:03:04 > 0:03:08technology for deaf people has come a long way in a relatively short
0:03:08 > 0:03:14space of time. Here at the Rochester School of the death, this museum and
0:03:14 > 0:03:19the items within it really highlight the rate of change of technology
0:03:19 > 0:03:22that aids communication for deaf and hard of hearing people, from things
0:03:22 > 0:03:27that look fairly primitive to us now, from ear trumpets to speaking
0:03:27 > 0:03:30tubes, to the cutting edge technology, such as live translation
0:03:30 > 0:03:35we're seeing today.The advancement of technology represents an
0:03:35 > 0:03:38opportunity for the advancement of deaf people in the workplace.RIT
0:03:38 > 0:03:43and in t.i.d. Started experiment in with speech recognition in the
0:03:43 > 0:03:47classroom so deaf and hard of hearing students who may not know
0:03:47 > 0:03:53sign language can participate fully. This lecture's sign language is
0:03:53 > 0:03:58being interpreted by a translator. It is then displayed on the board in
0:03:58 > 0:04:03real time.It will probably help me a lot. When the teacher is signing,
0:04:03 > 0:04:07I have an interpreter voicing form E, if I miss something while taking
0:04:07 > 0:04:13notes I can look at the screen and see what the teacher said. -- for
0:04:13 > 0:04:17me.One of the best things I've found on the laptop, when you can
0:04:17 > 0:04:22project it to your screen, you can see the closed captioning on your
0:04:22 > 0:04:26laptop. It allows you to save transcripts so you can use for
0:04:26 > 0:04:34future reference.Here at RIT we provide 25,000 hours of captioning
0:04:34 > 0:04:40every year for our deaf and hard of hearing students who are in classes
0:04:40 > 0:04:45primarily with hearing professor who is speaking their lecture. We
0:04:45 > 0:04:50thought one way we can cover those uncovered ours is to use automatic
0:04:50 > 0:04:56speech recognition. Captioning has been improved a lot in the past few
0:04:56 > 0:05:01months -- uncovered hours. It has improved last year when the error
0:05:01 > 0:05:10rate was so high. But now ASR has improved to the point where we feel
0:05:10 > 0:05:14very confident to pilot the programme to see how well it works.
0:05:14 > 0:05:19And we can also make improvements with the language model.One of the
0:05:19 > 0:05:22main challenges of speech recognition has been finding a way
0:05:22 > 0:05:26for deaf and hearing people to communicate in situations where
0:05:26 > 0:05:30there isn't an interpreter around. These students are using a special
0:05:30 > 0:05:34Instant Messenger app developed at the university. We all know how
0:05:34 > 0:05:41inaccurate speech recognition can be. -- special speech recognition
0:05:41 > 0:05:48app.I think automatic speech recognition is just getting started
0:05:48 > 0:05:52really and it's going to be improving. I think it opens up a
0:05:52 > 0:05:59whole new avenue that hasn't been possible before.We want to become
0:05:59 > 0:06:03the hub for really experimenting and looking for solutions to reduce the
0:06:03 > 0:06:09communication barriers that separate deaf and hearing people.
0:06:09 > 0:06:15That was Paul Carter. Now, in the UK, around 5% of all rail journeys
0:06:15 > 0:06:20are made by those with a disability or a long-term illness. That equates
0:06:20 > 0:06:26to around 85 million rail journeys every year. Now, while not all
0:06:26 > 0:06:30disabled people require assistance, a quarter have reported problems
0:06:30 > 0:06:34with using public transport. The rail company London Midland is
0:06:34 > 0:06:38hoping to improve accessibility for its De Zeeuw all passengers with a
0:06:38 > 0:06:43new app, Passenger Persist, and we asked Emily Yates to try it out for
0:06:43 > 0:06:50us. I'm Emily Yates and I'm just
0:06:50 > 0:06:55planning my train journey to Birmingham. It requires a fair bit
0:06:55 > 0:06:58of advanced booking. I'm confident travelling by myself but I'm not a
0:06:58 > 0:07:02huge fan of the train, which is actually why I'm making this
0:07:02 > 0:07:08journey. I've heard about an at in development called Passenger Assist
0:07:08 > 0:07:18which could be a game changer for disabled travellers. I think anybody
0:07:18 > 0:07:22watching this who's disabled will probably agree with me that you can
0:07:22 > 0:07:25have some pretty horrific travel journeys if you're disabled. I've
0:07:25 > 0:07:30been left on the train before, but persistence and somebody has said,
0:07:30 > 0:07:34yeah, we're going to come and meet you and I've been left on the train
0:07:34 > 0:07:38unable to get off and I've had to go four or five stops down the line to
0:07:38 > 0:07:43be able to come back again so I'm relieved as did to see what this app
0:07:43 > 0:07:56has to offer. -- really excited. I've got this new app which is
0:07:56 > 0:07:59currently in development and I'm just about to fill in my own
0:07:59 > 0:08:03profile. What's brilliant about this app is it says do you need room for
0:08:03 > 0:08:07a guide dog, do you have a hearing impairment, do you need a ramp, do
0:08:07 > 0:08:13you need help buying a ticket? So, Roxanne, I've added my profile
0:08:13 > 0:08:16details and now I've just planned a journey. This is obviously in
0:08:16 > 0:08:20development but this is how it would work. I've put in my journey and now
0:08:20 > 0:08:25it should come up on your phone any minute.Here you are...You've got
0:08:25 > 0:08:29my picture and everything so you know exactly what I'll look like.I
0:08:29 > 0:08:34know what you look like, know what to expect, I press I'm here to help.
0:08:34 > 0:08:41Brilliant.I can send you a message saying I'm here, my name is Roxanne.
0:08:41 > 0:08:46Back on the train for me and now I have this. Passenger Assist is being
0:08:46 > 0:08:53developed by start-up Transreport under the guidance of London
0:08:53 > 0:09:03Midlands' lab. Right now the phone is tracking both the staff member
0:09:03 > 0:09:07and the passenger. We're obviously in the same place so you can see the
0:09:07 > 0:09:11two dots are quite close together. Using the technology such as
0:09:11 > 0:09:17beacons, Wi-Fi, four G, GPS, we can use multiple tools.You know exactly
0:09:17 > 0:09:21what carriage I'm in even.Yes, down to the carriage, we can pinpoint
0:09:21 > 0:09:25less than one metre to your location and find out which carriage, which
0:09:25 > 0:09:34train and the direction of travel you're going in as well.That's all
0:09:34 > 0:09:39well and good for staff stations but how would it work on unmanned
0:09:39 > 0:09:42stations?The app is still in development, it's making good
0:09:42 > 0:09:46progress but we do have a challenge with unmanned stations. What we
0:09:46 > 0:09:50would want to do is get you to the Nehra is accessible station to get
0:09:50 > 0:09:55you on your way.And it's not just about the app. Transreport is also
0:09:55 > 0:09:58making trackable wristbands and these key faults, they'll be
0:09:58 > 0:10:02available for those unable to use phones as easily. Let's face it, in
0:10:02 > 0:10:07a world where we can now track our pizza by the minute am having to
0:10:07 > 0:10:11book train assistance 24 hours in advance seems a little old school
0:10:11 > 0:10:17and this way staff will have information at their fingertips too.
0:10:17 > 0:10:20Transreport hope to roll out the app across the London Midland service
0:10:20 > 0:10:25early next year and the plan is for the entire UK rail network to be
0:10:25 > 0:10:31able to access it by June, 2018. If there's one thing disabled
0:10:31 > 0:10:34travellers need that the current system doesn't provide its the
0:10:34 > 0:10:38reassurance that someone will be there to help and not leave them
0:10:38 > 0:10:48stranded when getting home all of the train. -- on all of the train.
0:10:48 > 0:10:53Hello and welcome to the week in tech. It was the week where the
0:10:53 > 0:10:57world's largest lithium ion battery was turned on in Australia.
0:10:57 > 0:11:01Assembled by Tesla and designed to store energy from a nearby wind
0:11:01 > 0:11:08farm, it can power 30,000 homes for one hour. An AI robot called Sophia
0:11:08 > 0:11:14started crowd fending for more brainpower. Instagram confirmed a
0:11:14 > 0:11:19pregnant Beyonce picture was the most liked 2017 so far and dogs were
0:11:19 > 0:11:24the most liked face filter. Speaking of selfies, Facebook's new security
0:11:24 > 0:11:28system may be asking for one to prove you're not a bot while logging
0:11:28 > 0:11:33on and Uber made headlines again, this time it was accused of using
0:11:33 > 0:11:38ex- CIA agents to spy on rivals also developing self driving tech. Plus
0:11:38 > 0:11:43it was revealed their data breach reported last week affected 57
0:11:43 > 0:11:47million users' details. Robots have been flexing their muscles, lifting
0:11:47 > 0:11:531000 times their own weight. These origami inspired bots from MIT mimic
0:11:53 > 0:11:58real muscles, being built with soft materials to make them a bit more
0:11:58 > 0:12:03human friendly. Universities across the world are working on these types
0:12:03 > 0:12:07of highly dextrous robots. And finally, good old-fashioned Lego is
0:12:07 > 0:12:12getting its own augmented reality app. The Lego AR studio will bring
0:12:12 > 0:12:19to life some of the most popular sets, so just as you thought the
0:12:19 > 0:12:22kids were sitting happily not having any screen time, well, out comes the
0:12:22 > 0:12:23app.
0:12:34 > 0:12:39I really began with my own reality. Someone using a wheelchair to get
0:12:39 > 0:12:43around and consistently frustrated when I show up at bases and I don't
0:12:43 > 0:12:47know which they are accessible or not until I get there. Countless
0:12:47 > 0:12:51times I show up and there are steps or other barriers that prevent me
0:12:51 > 0:12:57from doing the things that I want. And so I was really motivated to try
0:12:57 > 0:13:01and solve this problem and the way that we've gone about doing that is
0:13:01 > 0:13:05by creating a mobile app that can simply allow people who experienced
0:13:05 > 0:13:12express it -- accessibility needs to share information about what is
0:13:12 > 0:13:17accessible on their own communities and around the world. It starts by
0:13:17 > 0:13:20selecting a place and then breaking a place as accessible or not
0:13:20 > 0:13:25accessible. You can go one step further and and a description, you
0:13:25 > 0:13:29can say things like, I showed up at this place, the customer service was
0:13:29 > 0:13:34fantastic. So this cafe looks like it's not accessible. There are two
0:13:34 > 0:13:41steps at the entrance. But when I look at app my, I can see there's an
0:13:41 > 0:13:44alternative entrance through the building here, that will let me into
0:13:44 > 0:13:51the cafe. Let's check it out. For me the main magic, the most exciting
0:13:51 > 0:13:57part about Access, is the information is all crowd sourced
0:13:57 > 0:14:00from people who have experienced accessibility needs in there own
0:14:00 > 0:14:05life, or are just motivated to get wheelchair information. We started
0:14:05 > 0:14:10in Toronto with a couple of 100 pins and now we've reached over 20,000
0:14:10 > 0:14:14throughout the world. What we really want to make this a global movement.
0:14:14 > 0:14:20Before Access, I think people each had their own internal maps in their
0:14:20 > 0:14:25ahead of what places are accessible, what works for them in their own
0:14:25 > 0:14:28communities. But the second you step outside of your comfort zone of
0:14:28 > 0:14:31where you live in your neighbourhood, accessibility can
0:14:31 > 0:14:37become challenging. There are many times when people who have mobility
0:14:37 > 0:14:42needs are isolated in many ways and it simply because, you know, from
0:14:42 > 0:14:48the way I see it it's not people who are disabled but it's our
0:14:48 > 0:14:51environment that are disabling. So if we can remove the barriers that
0:14:51 > 0:14:56restrict people from engaging with their communities, with their
0:14:56 > 0:15:00workplaces, with their lifestyles, I think we can come to a much more
0:15:00 > 0:15:09inclusive world for everyone.That was from Toronto. Now, running a
0:15:09 > 0:15:16marathon, fancy it? No, me neither. But for Simon is a regular
0:15:16 > 0:15:21occurrence. We first met him a couple of years ago when he just ran
0:15:21 > 0:15:28and ultramarathon in Namibia. That's 150 miles. I should at this point
0:15:28 > 0:15:33say that Simon is blind, so knowing where to run is as much of a problem
0:15:33 > 0:15:38as the running itself. In Namibia he used audio from a smartphone to bite
0:15:38 > 0:15:44him. Since then, as a programmer and inventor, he's invented his own
0:15:44 > 0:15:49track to guide him while he runs. And the technology allowing
0:15:49 > 0:15:53navigation and visualisation of the world, with outside and in, is
0:15:53 > 0:15:56hopefully about to become a bit easier. We asked Simon to
0:15:56 > 0:16:03investigate.My name is Simon and I am a huge fan of technology. For the
0:16:03 > 0:16:07past few years I've really been using technology to push the
0:16:07 > 0:16:11boundaries of possibility. Recently I became the first blind person to
0:16:11 > 0:16:17attempt a marathon solo. A few weeks ago in New York I used technology to
0:16:17 > 0:16:22run a needed for 15 miles, until unfortunately the prototype didn't
0:16:22 > 0:16:29quite make it. For a blind person, mobility is always a key issue. Even
0:16:29 > 0:16:32things around the house, it is as simple as identifying different
0:16:32 > 0:16:36products in the kitchen or even identifying different articles of
0:16:36 > 0:16:42clothing, being able to identify objects is a constant problem. So
0:16:42 > 0:16:47one of the interesting breakthroughs has been a product from Microsoft
0:16:47 > 0:16:55which is and app that runs on an iPhone. Seeing Eye uses barcodes to
0:16:55 > 0:17:00recognise objects. You point the camera at the object. One of the
0:17:00 > 0:17:06really powerful things that it can do is simple text recognition.
0:17:06 > 0:17:13Vitamin D, wholegrain, serial, rice, sugar, salt flavouring...One thing
0:17:13 > 0:17:19is really nice is you can recognise people, so you can point your phone
0:17:19 > 0:17:27at somebody.In case you didn't catch that, that was Big Neil, seven
0:17:27 > 0:17:34feet away.One of the main issues is that AI algorithms usually identify
0:17:34 > 0:17:37objects in groups. What I really need is something to identify
0:17:37 > 0:17:43objects in detail, so don't just say a cup, identify it specifically and
0:17:43 > 0:17:50that's exactly what I've been able to work out. IBM are using their AI
0:17:50 > 0:17:52platform Watson to distinguish between similar objects. The
0:17:52 > 0:18:01prototype Big -- prototype app will be exciting if it works. First I had
0:18:01 > 0:18:06to do the training to get to grips with it.OK, when you are ready and
0:18:06 > 0:18:14you have your table lined with some objects, we can do some training.I
0:18:14 > 0:18:17had to take at least ten good pictures of an object against a
0:18:17 > 0:18:21solid background. Now with the model trains we are able to identify
0:18:21 > 0:18:28similar objects?Flower market.In the prototype stage I found the
0:18:28 > 0:18:33prototype really tricky. -- mug. I like that it was a voice guided so I
0:18:33 > 0:18:36could do it independently. I've got that connection with physical
0:18:36 > 0:18:42objects again. Photos and mugs and T-shirts. It got to the point where
0:18:42 > 0:18:45there was no distinctly -- distinction between them. Now if I
0:18:45 > 0:18:53train it and tag it I can get to know that individualisation again.
0:18:53 > 0:18:58It's great when you are doing things in the kitchen. But with certain
0:18:58 > 0:19:02objects which mean a loss to you it's not the same. To be able to tag
0:19:02 > 0:19:07them and find them it gives me that memory. That is really giving me
0:19:07 > 0:19:12something back that I lost many years ago. It leaves me really
0:19:12 > 0:19:23hopeful for the future.Simon... That was Simon and from his running
0:19:23 > 0:19:28guide to full on robotic limbs, we are seeing more and more tech
0:19:28 > 0:19:31advances that promise to help disabled athletes to compete in
0:19:31 > 0:19:35sports that were previously inaccessible to them. Our reporter
0:19:35 > 0:19:41had her legs amputated ten years ago and she's been trying out some new
0:19:41 > 0:19:46feet which could help amputees to carve up the slopes this winter.
0:19:46 > 0:19:52I've come to Dorset Orthopaedic a private company fits amputees with
0:19:52 > 0:19:57prosthetic legs, from running blades to hyperrealistic limbs. One sport
0:19:57 > 0:20:01that's always been very hard for me is snowboarding, because my normal
0:20:01 > 0:20:05feet designed specifically for walking. But here, they've got feet
0:20:05 > 0:20:09that would make that easier. Requirements of the full-time
0:20:09 > 0:20:12different with skiing compared to walking. With normal walking you
0:20:12 > 0:20:16need a foot that has a fairly small range of movement that gives you
0:20:16 > 0:20:21energy back, as you roll over the 40 get some push off at the end to help
0:20:21 > 0:20:25your walking. With skiing you need more movement in the foot to
0:20:25 > 0:20:30compensate for the uneven surface. You also need some shock absorption,
0:20:30 > 0:20:34so when you go over a bump or you land on the ski unique some that
0:20:34 > 0:20:39shock taken out and that's what this does.While I'm left to my own
0:20:39 > 0:20:43devices, Kevin agrees to fit my legs with some of these feet so I can
0:20:43 > 0:20:47give them a go. This requires a lot of honing and alignment to make sure
0:20:47 > 0:20:53I am not pushed too far forwards or backwards. Ta-da! My legs. I'm quite
0:20:53 > 0:20:59intrigued as to how this will feel. I have no idea.Oh, wow. If you push
0:20:59 > 0:21:03a weight forwards you should be able to feel the movement in the ankle.
0:21:03 > 0:21:07Oh, wow. That sensation I've not felt in the ten years since being an
0:21:07 > 0:21:12amputee. These feet work by putting your leg into an adjustable
0:21:12 > 0:21:16cylinder, which controls the resistance in the foot. More air and
0:21:16 > 0:21:19more resistance, as well as giving this movement, shock absorption in
0:21:19 > 0:21:24the foot means going over rocks or bumps is easier, they are not cheap.
0:21:24 > 0:21:28With the price of £2500 each, it means only some people can get
0:21:28 > 0:21:34access to them. It is clear that bear not for walking. They are very
0:21:34 > 0:21:39rigid, very square and very hard. But if I let my mind go and imagine
0:21:39 > 0:21:44myself snowboarding, which I've done badly in the past, I can feel that
0:21:44 > 0:21:50and then move and that's weird, because I have not felt my feet move
0:21:50 > 0:21:56in that way before. -- for ten years. There's only one thing left
0:21:56 > 0:22:01to do and it is to try defeat out on some proper snow. I've come to an
0:22:01 > 0:22:05indoor slope, but I've got to admit I'm feeling very nervous. This is
0:22:05 > 0:22:11Emma Gillespie. She has agreed to come with me to fit the feet and
0:22:11 > 0:22:18help the try them out. You've done it before.How was it? Hard.But you
0:22:18 > 0:22:26did it. So?One leg.This is what you don't see about being an
0:22:26 > 0:22:28amputee.
0:22:33 > 0:22:37When I've snowboarding previously on my normal walking feet, it's been
0:22:37 > 0:22:41really difficult. But these offer much more and the way they are set
0:22:41 > 0:22:45up office and natural bent on my knees, stands that almost impossible
0:22:45 > 0:22:58on neutral legs.Here we go. Think about your posture of it. -- a bit.
0:22:58 > 0:23:05And a turn!So it's been awhile since I boarded and expecting an
0:23:05 > 0:23:10instant result is probably asking a too much, but the best thing for it
0:23:10 > 0:23:14is to keep throwing myself down the slope and see what happens. Despite
0:23:14 > 0:23:18the technology of these feet, there is only so much they can do when it
0:23:18 > 0:23:22comes to hitting the slopes. The real work is definitely still coming
0:23:22 > 0:23:26from the person. And if you're not very good, bear not going to stop
0:23:26 > 0:23:36you from falling. -- they're. I'm soaking wet, that fall has drenched
0:23:36 > 0:23:42me. But it's amazing when you merge technology and disability, to give
0:23:42 > 0:23:46people independence and the feeling that they can try things that they
0:23:46 > 0:23:51perhaps thought weren't there for them. It's fun. But they are
0:23:51 > 0:24:00expensive. That's it for this special programme
0:24:00 > 0:24:06for the international day off People with Disabilities. We will carry on
0:24:06 > 0:24:10reporting on this stuff throughout the year, but you can keep a special
0:24:10 > 0:24:15eye on our disability stories online. Thanks for watching and see
0:24:15 > 0:24:16you soon.