:00:00. > :00:00.charts are compiled. But is it from me. We will be back at two o'clock
:00:00. > :00:08.but now here on BBC News it is time for Click.
:00:09. > :00:16.This week, watch out, pollution. We will clean up the city with a bird?
:00:17. > :00:47.No, a plane? No. It is a flying fish drone. This week is the BBC's so I
:00:48. > :00:51.can breathe season, looking at ways to tackle air pollution around the
:00:52. > :00:56.world. We are out on the streets of London to test a new camera from a
:00:57. > :01:01.thermal imaging company. It has a sensitivity to a range of gases
:01:02. > :01:05.which are invisible to the human night. The camera is supposed to be
:01:06. > :01:08.used by experts who know what they are looking for in numbers and
:01:09. > :01:12.colours that they see a bid is really supposed to be used in
:01:13. > :01:18.industrial locations as well where you are looking for gas leaks. But,
:01:19. > :01:22.I must say, even here I can see sprays coming from some of the
:01:23. > :01:33.exhaust pipes through this camera that I cannot see with my eyes. Now,
:01:34. > :01:36.if you want to tackle air pollution problems across a city, you have to
:01:37. > :01:40.know where the pollution is coming from and at what time of day. That
:01:41. > :01:49.is something that Mark has been investigating. Poorer quality, as a
:01:50. > :01:53.result of pollution, poses a serious risk to public health. It is a huge
:01:54. > :01:58.problem. The global burden of disease data now suggests that a
:01:59. > :02:01.lack of clean air is the third leading cause of death in the world
:02:02. > :02:06.after high blood pressure and smoking. But whether it triggers
:02:07. > :02:10.allergy or asthma, understanding the exact challenges that pollution
:02:11. > :02:19.causes, especially in a city, can be tricky. The levels of pollution in
:02:20. > :02:22.cities can vary a lot between individual streets. The more precise
:02:23. > :02:26.the information is, the better we can come up with strategies to
:02:27. > :02:33.improve things. We can identify areas with particular problems.
:02:34. > :02:36.Action to gather that even more precise data about pollution is
:02:37. > :02:42.being taken on the other side of the Atlantic, in Chicago. Because of the
:02:43. > :02:46.location of Chicago in the midwest and the fact that it is a large
:02:47. > :02:51.city, it is something of a transport hub for road, rail and air
:02:52. > :02:57.travellers. All those different vehicles don't do any favours for
:02:58. > :03:02.the air quality in the city. Here, a system is being installed which has
:03:03. > :03:06.been dubbed a fitness track for a city. It is called the array of
:03:07. > :03:12.things when it is completed it will be a citywide network of sensors
:03:13. > :03:16.fitted to lampposts and polls. The rate will monitor a variety of
:03:17. > :03:20.things from local climate to traffic levels and the air quality of the
:03:21. > :03:31.city. Eventually, all of the data the array gathers will be made
:03:32. > :03:37.available online for anybody to use. We have come just outside of Chicago
:03:38. > :03:42.to the Argonne National laboratory. It is part of the US Department of
:03:43. > :03:55.energy and is the birthplace of the array of things. The donor is really
:03:56. > :04:01.into air quality, so they are really excited. Here, the team behind the
:04:02. > :04:05.array continue to refine the centre boxes and the technology they can
:04:06. > :04:08.tame, blazing with city officials and arranging the continual roll-out
:04:09. > :04:14.of the network across the city. This is the guts, if you like, of the
:04:15. > :04:20.array of things. Which party you if the air quality sensor? This one is
:04:21. > :04:26.the air quality sensor. It is an elegiac cell at tuned to a Pacific
:04:27. > :04:30.type of chemical. This is a carbon monoxide one, this is the hydrogen.
:04:31. > :04:37.And it will record the total level of gas. Installation of the array
:04:38. > :04:41.began towards the end of 2016. By the end of 2018, 500 nodes are
:04:42. > :04:47.planned for the network, spread across different parts of the city.
:04:48. > :04:50.Charlie Kaplan is the project lead. He took me on a whistlestop tour of
:04:51. > :04:57.some of the city's earlier census sites. So, Charlie, this is the site
:04:58. > :05:03.of one of your first sensors, isn't it? This is one of the first six.
:05:04. > :05:08.This one here does the air quality, not just the general air quality but
:05:09. > :05:12.this one will tell us seven different gases and so that means we
:05:13. > :05:16.can say, well, this one is reading this gas particularly high and we
:05:17. > :05:21.know that that that is associated with a diesel truck. The new ones
:05:22. > :05:25.that we are putting in we have added an sensor for particles. What we can
:05:26. > :05:30.do with a particle sensor is we can look at the very fine particles that
:05:31. > :05:37.are measured by EPA and other organisations. The smaller particles
:05:38. > :05:40.are the ones you cannot see but they are quite dangerous. They will go
:05:41. > :05:43.straight into your bloodstream. The large ones are what triggers
:05:44. > :05:47.allergy. So if you are somebody with allergies related to asthma, you
:05:48. > :05:53.will be able to use the data from these nodes to look at Poland across
:05:54. > :05:58.the city and you may decide to change your cycle route to school or
:05:59. > :06:02.work, a sum may be where the pollen concentration is around the city.
:06:03. > :06:08.Chicago was not alone when it comes to pollution monitoring. We have a
:06:09. > :06:11.system also in London which combines historical pollution data with
:06:12. > :06:18.current pollution measurements to provide an hourly update of
:06:19. > :06:21.pollution levels across the city. The rollout in Chicago continues.
:06:22. > :06:25.The array of things nodes have been installed in other US cities with
:06:26. > :06:30.one in Seattle and another in Denver and there is interest in the city --
:06:31. > :06:35.system internationally as well. The data generated by the array of
:06:36. > :06:37.things will be used by researchers, scientists and healthcare
:06:38. > :06:42.professionals to get a better picture of the effects of poor air
:06:43. > :06:46.quality and pollution. When it comes to turning this information into
:06:47. > :06:56.action, that is job of local government. These two employees
:06:57. > :07:00.works of the city of Chicago and working out how the array of things
:07:01. > :07:06.can help city look at a range of issues. We have pockets of increased
:07:07. > :07:09.rates of asthma among our children that doctors have known about for
:07:10. > :07:14.quite sometime but they do not have a lot of information on why they
:07:15. > :07:20.happen in certain areas of the city. The role of the array is to help us
:07:21. > :07:24.understand the issues with air quality in Chicago in a detailed
:07:25. > :07:27.level because you cannot fix a problem if you cannot define it and
:07:28. > :07:33.understand it. We think about how heavy pollutant vehicles, say, if we
:07:34. > :07:39.installed hundreds of miles of biplanes, there is clear research
:07:40. > :07:43.showing that inhaling diesel fumes, especially by cyclists as they ride
:07:44. > :07:48.alongside traffic, can harm them. It helps us to picture and take a good
:07:49. > :07:53.look at where the bike avenues are and how that corresponds with the
:07:54. > :07:57.system. If you have a school or a vulnerable location close to an area
:07:58. > :08:01.that has increased air quality challengers, the data from the array
:08:02. > :08:05.of things will give us the ability to define a policy that will address
:08:06. > :08:10.that. A good example here in Chicago will be a quickly growing
:08:11. > :08:14.neighbourhood on the west side. It has evolved into one of our
:08:15. > :08:18.trendiest residential and entertainment district. But it has
:08:19. > :08:23.also been crisscrossed by any number of street level railroads. By
:08:24. > :08:27.looking at data and using data we will make decisions more confidently
:08:28. > :08:31.and we will know we'd better than many other cities have the ability
:08:32. > :08:35.to know that because of the data that we collect. Here, the
:08:36. > :08:42.technology has a role to play in the fight against poor air quality. But
:08:43. > :08:49.the bigger pollution busting powers relate to local and national
:08:50. > :08:56.government. That was market in Chicago. In London, I'm checking out
:08:57. > :09:02.a pollution monitoring device with a difference. I will give you a
:09:03. > :09:07.clue... This is the launch pad. With this water tank, they can launch
:09:08. > :09:18.their prototype. They even have their own in tunnel. Imperial
:09:19. > :09:22.College in London have a drone that can fly through the air, dive into
:09:23. > :09:28.the water and then leapt out again. Splash! All the while, gathering
:09:29. > :09:31.data to give us a greater understanding of pollution levels
:09:32. > :09:39.above and below the surface. The plan is to release a swarm of them
:09:40. > :09:44.into an area of concern. This is our response to extreme environments or
:09:45. > :09:49.post disaster applications such as after floods toxic spills, or I'll
:09:50. > :09:53.spills, nuclear accidents or so numb is. They are different classes of
:09:54. > :10:00.applications and capital abilities that they had to do something. This
:10:01. > :10:03.low-cost tool brings an enormous value compared to many other methods
:10:04. > :10:08.such as the human going there with a full protective suit. I was going to
:10:09. > :10:12.say, we have seen a lot of quite robots and we have seen a lot of
:10:13. > :10:17.flying robots. It never occurred to me that is quite difficult to get an
:10:18. > :10:23.underwater robot over great distances quickly and, so, you have
:10:24. > :10:28.combined the two. That is hard-core. So, yes, we will just die of it in
:10:29. > :10:34.the water and then died out and fly it that way. In some applications it
:10:35. > :10:39.is not even accessible through the water, in floods or ice you may not
:10:40. > :10:43.get there. On the other side, and aerial beacon may not be able to get
:10:44. > :10:49.the information that local people need so combining the two makes
:10:50. > :10:53.sense. During a dive, the drone fills with water and then by
:10:54. > :10:58.releasing carbon dioxide from its on-board gas chamber it forces the
:10:59. > :11:01.water back out as a high-powered jet which thrusts the drone back
:11:02. > :11:07.upwards, propelling it into the air. And in the wings unfold and it comes
:11:08. > :11:12.out of the water and it beautifully becomes this flying birdlike thing.
:11:13. > :11:18.It is quite graceful. That was a very romantic description. Now you
:11:19. > :11:22.know what sort of guy I am and what I get excited about. There is a
:11:23. > :11:27.beautiful part of it which makes it elegant. And elegance in nature that
:11:28. > :11:31.makes it effective as well. Having the folding wings might be beautiful
:11:32. > :11:36.but for us it allows us to reduce the drag that it would experience of
:11:37. > :11:42.the dives in the water and allows it to dive more deeply, as well is
:11:43. > :11:47.protecting the wings an impact. Hello and welcome to the week in
:11:48. > :11:52.Tech. A week which saw Airbus revealed plans for a hybrid car that
:11:53. > :11:57.flies. When Jaguar Land Rover revealed a search and rescue vehicle
:11:58. > :12:01.that is home to a heatseeking drain. And when high polluter showed off a
:12:02. > :12:06.500 metre long test tunnel from which it hopes to fire passengers at
:12:07. > :12:10.around 600 miles an hour. That would be a two second journey. Time to
:12:11. > :12:16.scream. Testing begins soon. It was also the week in which the
:12:17. > :12:21.revelation was televised. According to WikiLeaks, does the CIA can
:12:22. > :12:25.listen in on targets using Samsung TVs. Even when users think they have
:12:26. > :12:31.switched off. A range of other surveillance methods were exposed
:12:32. > :12:35.including a spy department dedicated to hacking the products of apple.
:12:36. > :12:40.WikiLeaks say that the CIA is out of control. Apple and Google say they
:12:41. > :12:43.have plug the holes and Samsung said it takes privacy seriously and will
:12:44. > :12:48.be listening closely to the concerns of its customer. Facebook was left
:12:49. > :12:52.red-faced when the BBC pointed out its platform was being used by
:12:53. > :12:55.convicted paedophiles to share sexualised images of children. And
:12:56. > :13:00.because the BBC shared the images with Facebook to help clean up its
:13:01. > :13:03.platform, Facebook reported the BBC to the police, accusing the
:13:04. > :13:09.corporation of distributing images of child exploitation. Want to buy a
:13:10. > :13:23.cheap house? This one took only 24 hour was to print and cost $10,000.
:13:24. > :13:26.artificial intelligence ahead of us, it is no surprise that tech giants
:13:27. > :13:31.are investing big time in data sensors. Super brains to make
:13:32. > :13:35.intelligent decisions in the cloud. But is this the best tactic? Here's
:13:36. > :13:44.Dave Lee. And video is taking a different
:13:45. > :13:47.approach. -- NVIDIA. It wants to do all that computation on this. NVIDIA
:13:48. > :13:50.is best known for creating chips to handle high-end graphics, at
:13:51. > :13:55.increasingly the company is looking to apply that computer power to data
:13:56. > :13:59.and AI. This week it introduced Jetson T X two, the latest in their
:14:00. > :14:06.line of what are essentially supercomputers on a chip. So, the
:14:07. > :14:10.Jetson TX2 is really for artificial intelligence at the age, devices
:14:11. > :14:13.like robots, drones, portable medical devices, which need a lot of
:14:14. > :14:17.intelligence, but they are really small and they have small power. So
:14:18. > :14:20.Jetson is going to give them the level of performance they need to do
:14:21. > :14:25.artificial intelligence in that small size. So a drone that has
:14:26. > :14:29.artificial intelligence on board is going to help find people that are
:14:30. > :14:35.missing in the wilderness, say, and find them and deliver them first aid
:14:36. > :14:38.and supplies. Experimenting with the new gear, people said it has many
:14:39. > :14:42.practical applications. There are many reasons why you might want to
:14:43. > :14:45.keep your computer power on a local device like this. For starters, it
:14:46. > :14:49.is much more secure, because your data is not being sent to and from
:14:50. > :14:53.the cloud constantly. That means some decisions are made quicker,
:14:54. > :14:58.which, if you are riding in a soft driving car, you will probably
:14:59. > :15:02.appreciate. -- self driving. There are many microcomputers on the
:15:03. > :15:06.market and most of them strive to be as cheap as possible. Not NVIDIA's.
:15:07. > :15:15.The Jetson TX2 will cost at least $400.
:15:16. > :15:23.It is that time of year again. I've arrived at London's wearable
:15:24. > :15:28.technology show. Only some of the highlights don't seem to actually be
:15:29. > :15:32.wearable. Well, I've always thought that one of the most natural uses
:15:33. > :15:38.for augmented reality would be to provide such now have in a car. --
:15:39. > :15:42.satnav. That is one of the functions this device provides. It has this
:15:43. > :15:45.section on the dashboard, when images reflected onto this small
:15:46. > :15:49.piece of glass, and then we also have this dial on the steering wheel
:15:50. > :15:52.which allows you to run through various functions. Things like being
:15:53. > :15:55.able to change or music, or answering phone calls without overt
:15:56. > :15:59.in your eyes away from that route straight ahead. The only thing is
:16:00. > :16:03.that you are actually changing the length of focus, so even though I'm
:16:04. > :16:06.looking in the same direction, looking at the screen does take my
:16:07. > :16:10.attention away from the road a little. Probably for less time than
:16:11. > :16:16.a separate satnav screen over there, though. Smart rings, vibrating
:16:17. > :16:23.coats, sportswear tracking your every move. It has all been thought
:16:24. > :16:31.of. The market for wearables reached an all-time high in 2016, with 102.4
:16:32. > :16:34.million devices shipped. But the focus has shifted away from smart
:16:35. > :16:40.devices connecting to multiple apps to simpler ones connecting to just
:16:41. > :16:44.one, and that seems to be a trend reflected here. If you are
:16:45. > :16:48.travelling somewhere on foot and you need to find your way, then some
:16:49. > :16:53.satnav in your shoes would of course be ideal. This device has been
:16:54. > :16:58.around a little while, which can attach to the laces of a pair of
:16:59. > :17:02.trainers. But now it also slips inside and insole, so if it is time
:17:03. > :17:06.to turn left, well, your left foot will vibrate. Time to turn right,
:17:07. > :17:10.and your right foot well. Last year we featured a different type of
:17:11. > :17:13.vibrating in Seoul. This is a prototype which is aimed at the
:17:14. > :17:18.elderly or firm to help them maintain balance. This year, the
:17:19. > :17:21.same company has a different product, a device for people
:17:22. > :17:25.suffering from Parkinson's. It will shine this laser light in front of
:17:26. > :17:30.each foot to help them put each but steadily in front of the other.
:17:31. > :17:37.Within Parkinson's there is a symptom called freezing of gate,
:17:38. > :17:41.which is fairly common. -- gait. It makes an individual feel as if they
:17:42. > :17:44.are glued to the floor at any moment during walking. As you can imagine,
:17:45. > :17:48.if your feet are suddenly not following you, you become quite
:17:49. > :17:53.prone to falling. Researchers found you can use visual triggers and
:17:54. > :17:58.sensory cues to enable a person to continue walking and taken another
:17:59. > :18:02.step. And another insole on display. This seems to be a theme this year.
:18:03. > :18:06.This time it is a personal safety alarm. If you want to activate it
:18:07. > :18:09.you type your feet together twice and you'll selected emergency
:18:10. > :18:13.contacts will be told there is an issue. -- your selected. To switch
:18:14. > :18:18.it off, you type your feet together three times. Some products on show
:18:19. > :18:22.were more finished than others, but overall it was a good glimpse at how
:18:23. > :18:29.some of the latest wearable tech is looking right now.
:18:30. > :18:33.That was Lara. Now, if you are a parent, like me, it has probably
:18:34. > :18:37.crossed your mind that your kids might be using technology a bit too
:18:38. > :18:42.much. How long are they spending on their phones? How much are they
:18:43. > :18:47.texting? But the popularity of texting amongst young people isn't
:18:48. > :18:52.all bad. Sumeet Dowson has been exploring how one organisation is
:18:53. > :18:58.using it to deal with serious issues for young people.
:18:59. > :19:04.Every Monday morning, this woman spends four hours texting with
:19:05. > :19:08.people in need. She is a volunteer counsellor for crisis text line, a
:19:09. > :19:15.free support service in the United States. Councillors and textures
:19:16. > :19:19.remain anonymous for privacy reasons. -- texters. We have a lot
:19:20. > :19:23.of middle schoolers who are concerned about what is going on and
:19:24. > :19:26.they reach out to us during the day. They might be concerned about
:19:27. > :19:29.sitting alone at lunch, for example. We have texters texting in because
:19:30. > :19:36.they are in a domestic violence situation. Most texters are young,
:19:37. > :19:40.under the age of 25. People tell us everything. They spill their guts.
:19:41. > :19:44.Typically by the third message. Nobody overhears you, you don't have
:19:45. > :19:48.to wait, even to be in a quiet place a quiet moment. The millions of
:19:49. > :19:53.messages exchanged on crisis text lines make up a data sets teeming
:19:54. > :19:56.with mental health insights. It reveals when texters struggle with
:19:57. > :20:02.eating disorders and where they have suicidal thoughts. The data was also
:20:03. > :20:06.used to build an algorithm. The model essentially performs triage by
:20:07. > :20:10.analysing each word in the message. So a person who is thinking about
:20:11. > :20:13.arming themselves would have a higher priority in the queue than
:20:14. > :20:18.somebody who is out after a breakup. -- harming. We quickly learned there
:20:19. > :20:27.were other things that were even more high-risk, that we didn't think
:20:28. > :20:30.of or didn't know. Things like #kms, which means "Kill myself".
:20:31. > :20:36.Conversations which reference things like either broken, Tylenol, and
:20:37. > :20:42.Bill, draino, all the household drugs that are within reach. The
:20:43. > :20:48.data is another mice and texters can opt out of data sharing. To promote
:20:49. > :20:51.mental health research some data is shared with researchers. Scientists
:20:52. > :20:55.at Stanford use natural language processing to study about 3 million
:20:56. > :20:59.text messages. They uncovered five phases in the conversations. The
:21:00. > :21:04.introduction, problem setting, exploration, problem-solving and the
:21:05. > :21:08.wrapup. The best councillors were really quick to get through this
:21:09. > :21:11.problem exploration phase. They were really good at getting to the heart
:21:12. > :21:15.of the issue to understand that, and they were quicker to move on in the
:21:16. > :21:19.conversation, which means that they then had more time to spend in this
:21:20. > :21:24.problem-solving phase. At the end of the chat, texters can rate their
:21:25. > :21:27.experience and the council. The researchers found that effective
:21:28. > :21:33.councillors avoided canned responses and able to shift the texter's
:21:34. > :21:36.outlook. We built an algorithm set that could measure different kinds
:21:37. > :21:40.of perspective change from talking, using lots of negative words, to
:21:41. > :21:44.talking about more positive words, to talk about how much you focus on
:21:45. > :21:48.the past versus the present and future, and how much you focus on
:21:49. > :21:52.versus other people. The next step is to create training tools for
:21:53. > :21:56.councillors, like real-time feedback on the conversation, and exploring
:21:57. > :22:02.the potential of a conversational agent. A robot. While data science
:22:03. > :22:06.and tech gets these self-professed data nerds that the crisis text line
:22:07. > :22:17.very excited, it will not use chat box. -- bots. Every messages read
:22:18. > :22:20.and reply to buy a human. We couldn't let you go without
:22:21. > :22:28.mentioning this mind controlled robot that responds really well to a
:22:29. > :22:31.certain thought. In collaboration with Boston University, MIT's
:22:32. > :22:34.computer science and artificial intelligence laboratory has
:22:35. > :22:39.published a system which allows human uses to correct a robot's
:22:40. > :22:42.mistakes by thought alone. It uses the signal we produce when we detect
:22:43. > :22:50.a mistake. It is called the error potential. The user wears an EEG cap
:22:51. > :22:53.and watch as the robot sought paint and wire into two bins. If they see
:22:54. > :22:58.the robot making a wrong choice, they simply think, that's wrong! The
:22:59. > :23:04.cap picks up that thought and the robot will correct its mistakes. We
:23:05. > :23:07.are interested in exploring the possibility of combining the
:23:08. > :23:14.potential -- error potential with other types of signals, which might
:23:15. > :23:18.be easily reliable. Even though these are baby steps there are
:23:19. > :23:25.tremendous applications that could happen in the home, on the factory
:23:26. > :23:28.line, or in the floors, so this technology can help support people
:23:29. > :23:34.in their daily activities, whether they are at work, at play, or in
:23:35. > :23:38.transportation. Pretty interesting stuff, although admittedly, I think
:23:39. > :23:43.that is still in the far future. So how about I tell you about something
:23:44. > :23:50.in the more immediate future? Next week, click is going to India. We
:23:51. > :23:54.will be travelling across the country to meet the people working
:23:55. > :24:02.hard to change lives, save lives, and maybe one day try out a new
:24:03. > :24:06.life. I can't wait. It is going to be brilliant. Join us on Twitter
:24:07. > :24:09.throughout the week for more techies and behind-the-scenes photos and we
:24:10. > :24:17.will see you next week in India. -- Tech news.