1 00:00:03,200 --> 00:00:06,640 You might not think about it very often, but we are living 2 00:00:06,640 --> 00:00:09,360 through a golden age of technology. 3 00:00:09,360 --> 00:00:14,360 Our homes, and our pockets, are full of inventions and gadgets 4 00:00:14,360 --> 00:00:17,360 that 20 years ago would have been the stuff of science fiction. 5 00:00:17,360 --> 00:00:20,480 Alexa, how exactly do you work? 6 00:00:20,480 --> 00:00:24,680 Mm, I don't know that. That's what we're going to find out. 7 00:00:24,680 --> 00:00:27,560 These things have become such an integral part of our everyday lives 8 00:00:27,560 --> 00:00:30,920 that we've forgotten to remember how clever they are. 9 00:00:30,920 --> 00:00:35,280 I'm Professor Hannah Fry, and I want to open your eyes... 10 00:00:35,280 --> 00:00:37,120 I'm having a great time! 11 00:00:37,120 --> 00:00:40,880 ..to the secret genius of everyday objects. 12 00:00:40,880 --> 00:00:43,920 The ones that we take for granted, 13 00:00:43,920 --> 00:00:47,440 items that owe their existence to generations 14 00:00:47,440 --> 00:00:49,520 of scientific innovation. 15 00:00:49,520 --> 00:00:53,200 THE Edison drove an electric car? Yeah. This one. 16 00:00:53,200 --> 00:00:58,320 In every episode, I'm going to take one item 17 00:00:58,320 --> 00:01:01,080 and tear it apart... 18 00:01:01,080 --> 00:01:03,600 ..bit by bit. 19 00:01:03,600 --> 00:01:06,400 So now there's some stuff that you can see. 20 00:01:06,400 --> 00:01:08,120 Meet the people behind it... 21 00:01:08,120 --> 00:01:10,240 Frankly, I thought many things were impossible. 22 00:01:10,240 --> 00:01:12,200 ..and unlock the amazing... 23 00:01:12,200 --> 00:01:15,160 Here is the clandestine listening device. 24 00:01:15,160 --> 00:01:19,280 ..and sometimes hilarious stories behind its invention. 25 00:01:19,280 --> 00:01:21,520 Hello! Look at you. 26 00:01:23,720 --> 00:01:28,720 This time, it's the food delivery app on your phone. 27 00:01:28,720 --> 00:01:33,320 It's almost become too easy to order delivery food. 28 00:01:33,320 --> 00:01:35,360 Whatever you like, whatever cuisine, 29 00:01:35,360 --> 00:01:40,360 you can just have it at your door in a matter of minutes 30 00:01:40,360 --> 00:01:42,760 just by pressing a few buttons on your phone. 31 00:01:42,760 --> 00:01:46,360 It's all thanks to some pensioners in Gateshead ordering Greggs 32 00:01:46,360 --> 00:01:47,960 in the 1980s. 33 00:01:47,960 --> 00:01:50,960 It's 20p! Oh, if only! 34 00:01:50,960 --> 00:01:53,520 'Pizza by internet'. 35 00:01:53,520 --> 00:01:56,320 Pizza pioneered online delivery. 36 00:01:56,320 --> 00:02:00,560 We actually got questions of, "Does the pizza come out of the floppy drive?" 37 00:02:01,760 --> 00:02:04,960 Oh, and submarine warfare and Soviet satellites. 38 00:02:04,960 --> 00:02:07,680 They made sure your takeaway doesn't get lost. 39 00:02:07,680 --> 00:02:11,160 'One of the great scientific feats of the age.' 40 00:02:21,520 --> 00:02:25,360 From Deliveroo, to Uber Eats... 41 00:02:25,360 --> 00:02:27,520 ..from Food Hub, to Just Eat, 42 00:02:27,520 --> 00:02:30,720 24 million of us now use apps 43 00:02:30,720 --> 00:02:34,240 to order an endless array of takeaways. 44 00:02:34,240 --> 00:02:39,320 Deliveroo, founded in 2013, has conquered every corner 45 00:02:39,320 --> 00:02:41,280 of the globe. 46 00:02:41,280 --> 00:02:44,440 And I'm going to one of their leading innovation hubs... 47 00:02:47,200 --> 00:02:49,720 ..in an industrial park in Acton. 48 00:02:52,520 --> 00:02:55,920 What has always been interesting to me is the burger 49 00:02:55,920 --> 00:02:58,800 is always the top selling item everywhere in the UK. 50 00:03:00,120 --> 00:03:03,760 This is the company's founder, Will Shu. 51 00:03:03,760 --> 00:03:06,200 Is that true in other countries? What about in Europe? 52 00:03:06,200 --> 00:03:08,760 Italy was not burgers. I think Italy was actually sushi, 53 00:03:08,760 --> 00:03:11,320 believe it or not. Really?! Yeah. Yeah. 54 00:03:13,200 --> 00:03:17,640 Whether it's delivering Mackie in Milan, or cheeseburgers in Chester, 55 00:03:17,640 --> 00:03:20,560 because Deliveroo's sole route to customers 56 00:03:20,560 --> 00:03:22,760 is via the delivery app, 57 00:03:22,760 --> 00:03:27,240 they can have kitchens cooking food exclusively for delivery. 58 00:03:27,240 --> 00:03:30,920 Think of it as a worldwide buffet that comes to you. 59 00:03:30,920 --> 00:03:32,320 Here you go. 60 00:03:32,320 --> 00:03:35,120 This thing is primed for efficiency, right? 61 00:03:35,120 --> 00:03:37,040 Because you don't have in-house dining. 62 00:03:37,040 --> 00:03:39,360 And so everything is geared towards delivery. 63 00:03:39,360 --> 00:03:41,680 It makes things much faster, more efficient. 64 00:03:44,760 --> 00:03:48,480 Online food delivery apps couldn't exist without some breathtakingly 65 00:03:48,480 --> 00:03:50,440 sophisticated technology. 66 00:03:51,960 --> 00:03:56,280 You can slice the app into four crucial pieces of tech. 67 00:03:56,280 --> 00:03:59,240 From the mapping system... 68 00:03:59,240 --> 00:04:01,960 ..to the delivery rider's display. 69 00:04:01,960 --> 00:04:06,320 And it all starts with the first thing that we see 70 00:04:06,320 --> 00:04:08,960 when we open up the app - the order screen. 71 00:04:12,760 --> 00:04:16,360 It might appear simple, but the order screen that you see 72 00:04:16,360 --> 00:04:18,760 is unique to you. 73 00:04:18,760 --> 00:04:21,840 Anyone who's ever ordered Deliveroo on their phone will know 74 00:04:21,840 --> 00:04:26,000 that, as you open up the app, it gives you a delicious array 75 00:04:26,000 --> 00:04:29,520 of potential options. Here's mine. It's nine o'clock in the morning. 76 00:04:29,520 --> 00:04:31,680 I've got pizza, pizza, 77 00:04:31,680 --> 00:04:36,040 um, some pizza, some crips, and, uh, some pizza. 78 00:04:36,040 --> 00:04:37,480 Delicious. 79 00:04:37,480 --> 00:04:40,320 But what you see is going to change depending on how you used 80 00:04:40,320 --> 00:04:45,240 the app before. So if I get Declan - this is our fancy director. 81 00:04:45,240 --> 00:04:49,520 We've got some dim sum, got some Asian noodles, 82 00:04:49,520 --> 00:04:51,680 sushi, and more dim sum. 83 00:04:51,680 --> 00:04:55,960 Eh, you really are part of the metropolitan elite, aren't you, Declan? I can see. 84 00:04:55,960 --> 00:04:58,960 Eh, Laura. Oh, hello! This is what I'm talking about. 85 00:04:58,960 --> 00:05:02,880 We've got Mexican chicken burgers, dripping with sauce. 86 00:05:02,880 --> 00:05:06,720 Lovely. Kebabs, burgers, more burgers! 87 00:05:06,720 --> 00:05:09,320 And a curry, which I think demonstrates 88 00:05:09,320 --> 00:05:12,720 that you're my kind of person. The point here is that what you see 89 00:05:12,720 --> 00:05:16,680 when you open the app is personalised to you, 90 00:05:16,680 --> 00:05:21,120 and the things that you previously have ordered on the app. 91 00:05:21,120 --> 00:05:23,040 Deliveroo is working on the assumption 92 00:05:23,040 --> 00:05:25,480 that if you've like something once, you're probably going 93 00:05:25,480 --> 00:05:28,640 to like it again, even if that includes kebabs at 9am, Laura. 94 00:05:31,760 --> 00:05:36,680 But personalisation isn't the only feature of the order screen. 95 00:05:36,680 --> 00:05:40,800 As soon as we place our order, it shows an estimated delivery time 96 00:05:40,800 --> 00:05:45,040 based on the track record of the kitchen cooking it. 97 00:05:45,040 --> 00:05:47,840 Back to Will. 98 00:05:47,840 --> 00:05:51,480 So, who comes up with the estimate for how long something's going to take to prepare? 99 00:05:51,480 --> 00:05:54,880 We do, based on millions and millions of data points, right? 100 00:05:54,880 --> 00:05:58,000 And so, take a brand like Nando's, we've worked with forever. 101 00:05:58,000 --> 00:06:01,480 We have a very, very good idea of how long it's going to take. 102 00:06:01,480 --> 00:06:04,400 Whereas if you're a newer restaurant, maybe in an area 103 00:06:04,400 --> 00:06:07,440 where we haven't been as long, it's going to take us a longer time 104 00:06:07,440 --> 00:06:09,800 for our machine learning models to figure that out. 105 00:06:09,800 --> 00:06:12,480 If you're saying, this is how long we expect it for you to make 106 00:06:12,480 --> 00:06:15,000 this dish, does that mean that it's like a target 107 00:06:15,000 --> 00:06:17,600 for the kitchen? Yeah, it's a target for the kitchen. 108 00:06:17,600 --> 00:06:20,840 And so, sometimes if they're really busy, it's going to be a little hard 109 00:06:20,840 --> 00:06:22,600 to hit it. We understand that. 110 00:06:22,600 --> 00:06:26,160 If the food's not ready in time, that restaurant 111 00:06:26,160 --> 00:06:28,760 will get bumped down the list of options on your order screen, 112 00:06:28,760 --> 00:06:31,400 which means that over time, they'll get fewer orders 113 00:06:31,400 --> 00:06:32,720 and less money. 114 00:06:35,840 --> 00:06:41,920 And that puts added pressure on the restaurant to hit its targets. 115 00:06:41,920 --> 00:06:45,760 But it means that the choices on your order screen are 116 00:06:45,760 --> 00:06:49,920 constantly updating, to improve average delivery times. 117 00:06:51,760 --> 00:06:54,280 This is, I have to be honest, more sophisticated 118 00:06:54,280 --> 00:06:56,200 than I was expecting. I think I sort of thought, 119 00:06:56,200 --> 00:06:58,880 "Oh," you know, "there's an order, send someone." 120 00:06:58,880 --> 00:07:01,600 I thought it was a bit like... I mean, you know, in a way, 121 00:07:01,600 --> 00:07:03,960 it's pretty easy to sort of like, conceptualise it. 122 00:07:03,960 --> 00:07:06,600 But what's behind all of it is actually very complex. 123 00:07:06,600 --> 00:07:09,840 There's just a lot of maths behind it, basically. 124 00:07:09,840 --> 00:07:13,600 You're talking my language. And I know that's what you're into, so... 125 00:07:14,720 --> 00:07:17,760 Once again, we've got the mathematicians to thank. 126 00:07:17,760 --> 00:07:19,880 You're welcome. 127 00:07:19,880 --> 00:07:23,560 Hiding behind the scenes in every food delivery app 128 00:07:23,560 --> 00:07:28,000 are a series of deliciously sophisticated algorithms. 129 00:07:28,000 --> 00:07:30,400 They might be invisible to you, 130 00:07:30,400 --> 00:07:34,200 but it's these programmes that offer up a variety of dishes 131 00:07:34,200 --> 00:07:38,400 and ordering options that we now take for granted. 132 00:07:38,400 --> 00:07:42,960 They keep track of every customer's previous orders and every kitchen's 133 00:07:42,960 --> 00:07:48,000 cooking time, and translate that into a really simple to use system. 134 00:07:49,720 --> 00:07:52,480 You have the magic going on behind the scenes, 135 00:07:52,480 --> 00:07:56,800 while all you have to do is press a few buttons and seamlessly 136 00:07:56,800 --> 00:07:59,440 have your food appear at your door. 137 00:08:00,840 --> 00:08:04,360 The algorithms powering the order screen on our smartphones 138 00:08:04,360 --> 00:08:07,480 are just the latest in a long tradition of technologies 139 00:08:07,480 --> 00:08:11,320 that we use to get out of cooking dinner. 140 00:08:11,320 --> 00:08:14,880 It all began just over 100 years ago. 141 00:08:14,880 --> 00:08:19,240 The earliest known reference to takeaway ordering technology comes 142 00:08:19,240 --> 00:08:22,200 from a Chinese restaurant in Los Angeles. 143 00:08:23,920 --> 00:08:28,760 In 1922, Kin-Chu Cafe encouraged diners to place their orders 144 00:08:28,760 --> 00:08:32,520 between 11am and one in the morning 145 00:08:32,520 --> 00:08:35,720 - obviously, the perfect time for a chow mein - 146 00:08:35,720 --> 00:08:38,640 by contacting the restaurant using a piece of technology 147 00:08:38,640 --> 00:08:40,960 that had recently taken off in the area. 148 00:08:42,840 --> 00:08:44,800 The telephone. 149 00:08:44,800 --> 00:08:46,720 It was a brilliant idea. 150 00:08:46,720 --> 00:08:49,760 Los Angeles had the largest telephone switchboard in the world 151 00:08:49,760 --> 00:08:55,480 at the time, and so prawn toast by phone was a hit with Angelinos. 152 00:08:55,480 --> 00:08:59,400 Phone ordering transformed customer waiting times, because now 153 00:08:59,400 --> 00:09:04,080 they could order takeaway from home and collect it at their convenience. 154 00:09:07,080 --> 00:09:10,160 So when do you think that food ordering began its move 155 00:09:10,160 --> 00:09:12,960 from old school telephone to a screen? 156 00:09:18,440 --> 00:09:21,520 It was actually about 40 years ago. 157 00:09:21,520 --> 00:09:24,640 And even more surprisingly, the pioneers of this 158 00:09:24,640 --> 00:09:28,600 new online technology were pensioners from Gateshead. 159 00:09:30,800 --> 00:09:32,480 You can get, 160 00:09:32,480 --> 00:09:35,440 there's also a chemist, which is in the scheme, 161 00:09:35,440 --> 00:09:38,720 and also, a bakery. And you can order them as well. 162 00:09:38,720 --> 00:09:43,680 It's 1984, and the marvellous Mrs Jane Snowball 163 00:09:43,680 --> 00:09:46,240 is doing something revolutionary from the comfort 164 00:09:46,240 --> 00:09:48,480 of her own living room. 165 00:09:48,480 --> 00:09:51,000 I think it's wonderful. 166 00:09:51,000 --> 00:09:53,720 The remarkable thing about the order that Jane is placing 167 00:09:53,720 --> 00:09:56,520 is that she's not using the Internet to do it. 168 00:09:56,520 --> 00:10:01,120 This is at least five years before the World Wide Web came along. 169 00:10:01,120 --> 00:10:06,280 Let me give you a nice wave of nostalgia by talking about Teletext. 170 00:10:06,280 --> 00:10:10,320 Of course, I am far too young to remember Teletext properly, 171 00:10:10,320 --> 00:10:12,080 but you might. 172 00:10:12,080 --> 00:10:14,840 And you could use it to get information from your TV, 173 00:10:14,840 --> 00:10:18,240 things like the weather or the news. 174 00:10:18,240 --> 00:10:21,680 There was another version called Video Text, where it wasn't 175 00:10:21,680 --> 00:10:23,800 just about broadcasting information. 176 00:10:23,800 --> 00:10:27,800 You could effectively talk back to it, using your remote. 177 00:10:27,800 --> 00:10:29,720 Let me show you how it might have worked. 178 00:10:29,720 --> 00:10:31,520 So, if I go to the shopping page here... 179 00:10:31,520 --> 00:10:33,280 Oh, here we go! 180 00:10:33,280 --> 00:10:35,440 Look at this. Oh! 181 00:10:35,440 --> 00:10:37,000 It's like going back in time. 182 00:10:37,000 --> 00:10:39,560 Eh, right, so let's go for Greggs, number one. 183 00:10:39,560 --> 00:10:42,040 There we go. Feeling quite peckish. 184 00:10:42,040 --> 00:10:44,920 Um, I think I'm going to try and find the sausage rolls. 185 00:10:44,920 --> 00:10:48,360 So let's go for zero two for hot food, zero two. 186 00:10:48,360 --> 00:10:52,400 OK. Hot food. I mean, this is a retro delight, isn't it? 187 00:10:52,400 --> 00:10:57,400 OK. We have got sausage roll right at the top, zero one. 188 00:10:57,400 --> 00:11:00,160 Oh! OK. So it's got it. I've got like, 189 00:11:00,160 --> 00:11:02,880 my own little running receipt going on. 190 00:11:02,880 --> 00:11:04,720 One there and it's 20p! 191 00:11:04,720 --> 00:11:06,600 Oh, if only! 192 00:11:06,600 --> 00:11:08,760 If only that bit was still true. 193 00:11:08,760 --> 00:11:11,640 Confirm order number one. There we go. 194 00:11:11,640 --> 00:11:14,480 'Thank you for shopping'! 195 00:11:14,480 --> 00:11:16,280 SHE SIGHS 196 00:11:19,120 --> 00:11:22,600 Jane Snowball and her friends managed to get so far ahead 197 00:11:22,600 --> 00:11:27,160 of the curve, thanks to an invention by technology entrepreneur 198 00:11:27,160 --> 00:11:31,000 Michael Aldrich. 199 00:11:31,000 --> 00:11:34,800 Michael Aldrich's kind of special insight was being able to see 200 00:11:34,800 --> 00:11:38,200 that things that were being used for delivering information 201 00:11:38,200 --> 00:11:42,080 on the television, could be used to send requests 202 00:11:42,080 --> 00:11:43,840 for other information. 203 00:11:43,840 --> 00:11:48,280 That is Asha Rospigliosi, an e-commerce expert who worked 204 00:11:48,280 --> 00:11:50,480 with Michael back in the 1980s. 205 00:11:50,480 --> 00:11:53,640 Where did the inspiration for this come from? 206 00:11:53,640 --> 00:11:56,280 I mean, that's quite a leap, isn't it? As he told it to me, 207 00:11:56,280 --> 00:12:00,240 he was out walking his dog in the Sussex countryside with his wife, 208 00:12:00,240 --> 00:12:03,560 and she said, "We need to allow time to go to the supermarket, 209 00:12:03,560 --> 00:12:07,040 "pick up the weekend shop." And he said, "That's a bit of a bore. 210 00:12:07,040 --> 00:12:10,920 "Wouldn't it be nice if...?" And at the same time there was work 211 00:12:10,920 --> 00:12:15,640 going on in Newcastle about the impact of out-of-town superstores 212 00:12:15,640 --> 00:12:19,320 and how that was disadvantaging some people. 213 00:12:20,560 --> 00:12:23,520 Aldrich thought if pensioners couldn't get to the shops, 214 00:12:23,520 --> 00:12:26,120 could he bring the shops to them? 215 00:12:27,600 --> 00:12:31,960 So to do that, he invented a two way interactive system, 216 00:12:31,960 --> 00:12:34,400 using their television screens. 217 00:12:36,560 --> 00:12:39,840 What was different was the TV was connected by a telephone line 218 00:12:39,840 --> 00:12:42,240 to the computer in the supermarket. 219 00:12:42,240 --> 00:12:44,440 It was two way? It was two way. 220 00:12:44,440 --> 00:12:47,000 Not just displaying, but receiving too? Yes. 221 00:12:47,000 --> 00:12:50,360 And that was the vital difference that let people place orders 222 00:12:50,360 --> 00:12:51,560 for their food. 223 00:12:53,120 --> 00:12:57,120 It was a very clever idea, but the order screen was slow 224 00:12:57,120 --> 00:12:59,000 and it was fiddly to use. 225 00:13:01,960 --> 00:13:05,440 If this was ever going to work, the food order screen needed 226 00:13:05,440 --> 00:13:08,200 to be much more user friendly. 227 00:13:08,200 --> 00:13:12,200 And then a few years later, a long way away from Gateshead, 228 00:13:12,200 --> 00:13:16,000 in the research lab Cern in Switzerland, 229 00:13:16,000 --> 00:13:20,400 a man called Tim came up with the idea of the World Wide Web. 230 00:13:23,240 --> 00:13:26,880 Little did Tim Berners-Lee know that one of the first things 231 00:13:26,880 --> 00:13:30,880 his invention would be used to sell was pizza. 232 00:13:30,880 --> 00:13:33,880 We actually got questions of, "Does the pizza come out 233 00:13:33,880 --> 00:13:36,080 "of the floppy drive?" 234 00:13:36,080 --> 00:13:38,040 That is Steph Marr, 235 00:13:38,040 --> 00:13:41,040 and this is Doug Michaels. 236 00:13:41,040 --> 00:13:44,080 They were probably the first people to sell anything 237 00:13:44,080 --> 00:13:47,280 on the World Wide Web, using their imaginatively named website 238 00:13:47,280 --> 00:13:49,040 Pizzanet. 239 00:13:49,040 --> 00:13:52,080 In the early 1990s, they were running a software company 240 00:13:52,080 --> 00:13:55,000 helping businesses with their inventory systems. 241 00:13:55,000 --> 00:13:57,520 Pizza Hut was a major customer of ours. 242 00:13:57,520 --> 00:14:00,480 They had our systems in every one of their stores. 243 00:14:00,480 --> 00:14:03,520 In the manager's office there was a server 244 00:14:03,520 --> 00:14:06,120 that kept track of all their inventory. 245 00:14:06,120 --> 00:14:09,680 How much flour do we have? How much pepperoni do we have? 246 00:14:10,760 --> 00:14:14,400 Unlike most companies at the time, the computers in their system 247 00:14:14,400 --> 00:14:17,960 were all linked together, so they could access any store 248 00:14:17,960 --> 00:14:20,120 in the Pizza Hut chain. 249 00:14:20,120 --> 00:14:25,040 What we figured out is we could call in to the store 250 00:14:25,040 --> 00:14:28,720 through the inventory management system, and pretend to be 251 00:14:28,720 --> 00:14:33,960 one of their cash registers, and inject an order into the system 252 00:14:33,960 --> 00:14:38,400 like they had a mysterious additional cashier. 253 00:14:38,400 --> 00:14:41,800 Steph and Doug could send their orders to any store 254 00:14:41,800 --> 00:14:45,040 and get a response almost instantly. 255 00:14:45,040 --> 00:14:48,600 And that was the inspiration they needed to create probably 256 00:14:48,600 --> 00:14:51,720 the world's first Web-based order system. 257 00:14:51,720 --> 00:14:53,560 This was the front page. 258 00:14:53,560 --> 00:14:57,600 It asked for your name, your street address and a phone number. 259 00:14:57,600 --> 00:15:00,320 Did you do want thin crust or thick crust? Did you want extra cheese? 260 00:15:00,320 --> 00:15:03,640 Did you want mushrooms, did you want sausage? Who knows? 261 00:15:03,640 --> 00:15:05,920 All of the above, please. 262 00:15:05,920 --> 00:15:10,040 I just click, click on any - you know, how many pizzas I want. 263 00:15:10,040 --> 00:15:13,200 You heard right - pizza by internet. 264 00:15:13,200 --> 00:15:16,320 Thanks to Pizza Hut, a division of PepsiCo Incorporated. 265 00:15:16,320 --> 00:15:19,880 We hold a press conference and we were amazed it got picked up 266 00:15:19,880 --> 00:15:24,960 on national news broadcasts, because it was finally a way 267 00:15:24,960 --> 00:15:28,840 to communicate what the Internet was and how it could work. 268 00:15:28,840 --> 00:15:32,680 Pizza Hut is testing the order by Internet system in the university 269 00:15:32,680 --> 00:15:36,600 town of Santa Cruz, California, said to be a hotbed of Internet junkies. 270 00:15:36,600 --> 00:15:41,040 If the idea clicks in California, you can just bet it'll take off 271 00:15:41,040 --> 00:15:43,680 down the information superhighway. 272 00:15:45,200 --> 00:15:47,880 And the idea did well and truly take off 273 00:15:47,880 --> 00:15:52,120 down the information superhighway. With the roll-out of smartphones 274 00:15:52,120 --> 00:15:56,280 in the late 2000s, giving us an internet-enabled order screen 275 00:15:56,280 --> 00:15:58,880 right there in our pockets. 276 00:15:58,880 --> 00:16:03,200 But of course, placing your order is just the start. 277 00:16:06,240 --> 00:16:10,120 It's only worth ordering a cheeky tikka bhuna 278 00:16:10,120 --> 00:16:13,120 if you know that it's going to turn up at the right address. 279 00:16:13,120 --> 00:16:16,840 So here is our second piece of tech, 280 00:16:16,840 --> 00:16:19,680 the mapping technology that the riders depend on 281 00:16:19,680 --> 00:16:24,960 to find their way, using signals from satellites circling the Earth. 282 00:16:24,960 --> 00:16:27,160 It's GPS. 283 00:16:29,680 --> 00:16:34,280 And these kind folks are going to explain how GPS works. 284 00:16:34,280 --> 00:16:36,880 Pre-fast 45. Go 45. 285 00:16:36,880 --> 00:16:39,960 We're at Diego Stream Bravo. I'm SS02. 286 00:16:39,960 --> 00:16:43,520 No time tags, no cif TPs, we're in the window. Good six. 287 00:16:43,520 --> 00:16:46,960 Your IDs are good. Good seven. It's great. Diego on the Bravo 288 00:16:46,960 --> 00:16:49,960 you have both comm ones. Great. No cif TPs, no time tags, good nine. 289 00:16:49,960 --> 00:16:52,320 What just happened? 290 00:16:52,320 --> 00:16:55,480 Each position has a specific list of things that they check 291 00:16:55,480 --> 00:16:58,640 to make sure that everything is as it should be before we make contact 292 00:16:58,640 --> 00:17:00,400 and go live on a satellite. 293 00:17:00,400 --> 00:17:04,480 Given that you're running GPS for the entire world in this room, 294 00:17:04,480 --> 00:17:07,800 there's not that many people here. I was sort of expecting more. 295 00:17:07,800 --> 00:17:12,120 Yep. That is a common thing that people are shocked by. 296 00:17:12,120 --> 00:17:15,360 That is Lieutenant Sydney Lamond, 297 00:17:15,360 --> 00:17:19,240 and this is the Schriever Space Force Base, 298 00:17:19,240 --> 00:17:21,600 where worldwide GPS is controlled. 299 00:17:23,160 --> 00:17:26,080 From this building. Sydney's team controls 300 00:17:26,080 --> 00:17:30,480 the 31 satellites, 12,000 miles above our heads, 301 00:17:30,480 --> 00:17:32,240 that make up GPS. 302 00:17:34,360 --> 00:17:38,120 Sort of like air traffic control on steroids. 303 00:17:38,120 --> 00:17:40,560 Talk me through that then. So on a normal day, 304 00:17:40,560 --> 00:17:42,640 what kind of things are you doing in here? 305 00:17:42,640 --> 00:17:45,720 We generate the commands. Tell the satellite what to do from here. 306 00:17:45,720 --> 00:17:48,040 That information gets sent up to the satellite 307 00:17:48,040 --> 00:17:50,840 via a ground antenna. Satellite takes that in, 308 00:17:50,840 --> 00:17:53,320 does what it's supposed to, and beams down its new signal 309 00:17:53,320 --> 00:17:55,440 to our six billion users worldwide. 310 00:17:55,440 --> 00:17:57,960 OK. How - I mean...! 311 00:17:57,960 --> 00:18:00,960 It sounds like quite a lot for ten people to do. 312 00:18:00,960 --> 00:18:03,920 The satellites are spread out across the planet. 313 00:18:03,920 --> 00:18:08,080 So no matter where you are, you are almost always covered. 314 00:18:08,080 --> 00:18:13,760 They are set in set positions in six orbits around the Earth. 315 00:18:13,760 --> 00:18:16,440 OK, so it's like you have these kind of elliptical bands, 316 00:18:16,440 --> 00:18:18,040 that coat around. 317 00:18:18,040 --> 00:18:21,120 I'm sort of imagining like taking a watermelon and coating it 318 00:18:21,120 --> 00:18:23,760 in elastic. Yes. A lot of elastic bands, correct. 319 00:18:23,760 --> 00:18:27,000 Of course you've got it as a hologram. Of course you do. 320 00:18:27,000 --> 00:18:30,520 But the satellites don't automatically know their location 321 00:18:30,520 --> 00:18:32,480 in orbit around the Earth. 322 00:18:32,480 --> 00:18:37,120 Sydney's team have to tell the satellites where they are. 323 00:18:37,120 --> 00:18:40,120 So we tell it where it is, so that it can tell you guys 324 00:18:40,120 --> 00:18:43,640 where you are on the ground. So, wait - how do you know where it is? 325 00:18:43,640 --> 00:18:46,360 We have a huge algorithm in the background that runs 326 00:18:46,360 --> 00:18:50,400 all kinds of data and crunches numbers to essentially 327 00:18:50,400 --> 00:18:52,280 compute the GPS constellation. 328 00:18:52,280 --> 00:18:55,120 So essentially, in this room then, all these people, every day, 329 00:18:55,120 --> 00:18:57,600 you're telling the satellites where they are so they can tell 330 00:18:57,600 --> 00:19:01,120 everyone else in the world where they are. That is correct. Very good explanation. 331 00:19:01,120 --> 00:19:03,880 Well, it's a system I'm happy to use. Thank you for all the years 332 00:19:03,880 --> 00:19:06,880 when I haven't got lost. I appreciate that. You're welcome. 333 00:19:06,880 --> 00:19:10,680 But how do those satellites help the poor pizza delivery person 334 00:19:10,680 --> 00:19:15,000 cycling around South London, trying to find my house? 335 00:19:15,000 --> 00:19:19,080 So you've got all of these satellites floating around in orbit. 336 00:19:19,080 --> 00:19:23,360 How on earth do they work out your position? 337 00:19:23,360 --> 00:19:25,920 I know what you're thinking. "What is she doing, 338 00:19:25,920 --> 00:19:28,240 "rooting around in the bushes?" But don't worry. 339 00:19:28,240 --> 00:19:30,920 This will all make sense, I promise. 340 00:19:30,920 --> 00:19:33,920 I'm going to need a stick... 341 00:19:33,920 --> 00:19:36,560 ..like this one... 342 00:19:36,560 --> 00:19:40,720 ..to show you a process called trilateration. 343 00:19:40,720 --> 00:19:42,800 Come this way. 344 00:19:42,800 --> 00:19:46,080 OK, imagine that you've just ordered a pizza and the delivery guy 345 00:19:46,080 --> 00:19:48,720 is out there on his bike, he's got his phone with him. 346 00:19:48,720 --> 00:19:51,160 He's somewhere on the surface of the planet Earth. 347 00:19:51,160 --> 00:19:54,920 But you're not quite sure where. How do you work out his position? 348 00:19:54,920 --> 00:19:58,120 So, for this, I want you to imagine that I am a satellite 349 00:19:58,120 --> 00:20:02,960 and I'm sending out these signals at regular intervals all around me. 350 00:20:02,960 --> 00:20:07,440 When the GPS receiver in the rider's phone detects those signals, 351 00:20:07,440 --> 00:20:10,920 it can work out how far away it is from the satellite 352 00:20:10,920 --> 00:20:12,400 all the way up in space. 353 00:20:14,240 --> 00:20:17,240 And then, just from that, you can start to work out the rider's 354 00:20:17,240 --> 00:20:20,920 location because every possible point on Earth that is a fixed 355 00:20:20,920 --> 00:20:26,120 distance away from the satellite makes a circle, which I'm drawing 356 00:20:26,120 --> 00:20:27,960 with my trusty stick. 357 00:20:29,600 --> 00:20:34,320 I know that this delivery driver has to be lying somewhere 358 00:20:34,320 --> 00:20:37,520 on this line, somewhere on this circle. 359 00:20:37,520 --> 00:20:40,480 And that narrows it down quite nicely, but not enough 360 00:20:40,480 --> 00:20:43,280 to know his exact position, which is why you need 361 00:20:43,280 --> 00:20:44,640 more than one satellite. 362 00:20:44,640 --> 00:20:46,840 So now I'm a new satellite over here, 363 00:20:46,840 --> 00:20:49,720 also sending out signals, one of which is also picked up 364 00:20:49,720 --> 00:20:51,440 by the delivery driver's phone. 365 00:20:51,440 --> 00:20:57,080 And now, using the same logic, we know that the delivery guy 366 00:20:57,080 --> 00:21:01,520 has to be somewhere on this line, but also has to be somewhere 367 00:21:01,520 --> 00:21:06,160 on THIS line, which means you can deduce that where these two 368 00:21:06,160 --> 00:21:10,200 lines intersect, at that point or that point, that is where 369 00:21:10,200 --> 00:21:12,120 the delivery driver must be. 370 00:21:12,120 --> 00:21:15,680 So, with just two satellites, you can work out 371 00:21:15,680 --> 00:21:19,560 the rider must be in one of two possible places. 372 00:21:19,560 --> 00:21:22,840 Now, that's still not quite enough to narrow it down completely, 373 00:21:22,840 --> 00:21:27,440 but if you add in a third satellite, a third fixed distance, 374 00:21:27,440 --> 00:21:32,480 using that same logic, there is now only one possible place 375 00:21:32,480 --> 00:21:36,160 that the delivery guy can be, and that is where all three 376 00:21:36,160 --> 00:21:39,120 of those circles intersect, right in the middle 377 00:21:39,120 --> 00:21:41,760 and just outside your front door. 378 00:21:43,080 --> 00:21:46,040 You actually need four satellites to pinpoint precisely, 379 00:21:46,040 --> 00:21:48,400 but that's not the point here. 380 00:21:48,400 --> 00:21:53,160 Us pizza lovers have been relying on this tech since the early 2000s. 381 00:21:53,160 --> 00:21:56,680 The US military's love affair with GPS, however, 382 00:21:56,680 --> 00:21:59,000 that goes back much further. 383 00:22:04,560 --> 00:22:09,360 In fact, its origins go all the way back to the Cold War. 384 00:22:11,200 --> 00:22:15,960 At that time, there was a SLIGHTLY bigger worry than the pizza guy 385 00:22:15,960 --> 00:22:17,880 not finding your house. 386 00:22:17,880 --> 00:22:20,600 The submarine force is the largest in the world, 387 00:22:20,600 --> 00:22:23,080 approximately 350 undersea craft. 388 00:22:24,400 --> 00:22:25,960 Let me set the scene. 389 00:22:25,960 --> 00:22:29,960 It's the 1950s and tensions between the Soviets and the Americans 390 00:22:29,960 --> 00:22:32,000 have reached fever pitch. 391 00:22:35,280 --> 00:22:39,960 Nuclear weapons are now a reality, and the US are desperately trying 392 00:22:39,960 --> 00:22:44,040 to develop a navigation system so that their own nuclear submarines 393 00:22:44,040 --> 00:22:47,200 can know precisely where they are when they're out at sea, 394 00:22:47,200 --> 00:22:49,360 far away from any landmarks. 395 00:22:51,640 --> 00:22:55,720 At the same time, the two superpowers are also locked 396 00:22:55,720 --> 00:22:58,520 in a race to see who will be the first to conquer 397 00:22:58,520 --> 00:23:00,320 the new frontier - 398 00:23:00,320 --> 00:23:02,120 space. 399 00:23:03,400 --> 00:23:07,520 Ironically, a major breakthrough for the US military's navigation 400 00:23:07,520 --> 00:23:10,920 problem would come from a Soviet space triumph. 401 00:23:13,120 --> 00:23:17,720 ARCHIVE: Today, a new moon is in the sky, a 23-inch metal sphere placed 402 00:23:17,720 --> 00:23:20,120 in orbit by a Russian rocket. 403 00:23:20,120 --> 00:23:24,280 Sputnik, the world's first human-made satellite, was launched 404 00:23:24,280 --> 00:23:28,000 by the Soviet Union on 4th October in 1957. 405 00:23:28,000 --> 00:23:29,560 BEEPING 406 00:23:29,560 --> 00:23:33,480 You are hearing the actual signals transmitted by the Earth-circling 407 00:23:33,480 --> 00:23:37,560 satellite, one of the great scientific feats of the age. 408 00:23:38,920 --> 00:23:43,960 The Americans feared the Soviets would put Sputnik to sinister uses. 409 00:23:43,960 --> 00:23:47,560 The US know that Sputnik has been launched. 410 00:23:47,560 --> 00:23:49,360 No-one knows where it is. 411 00:23:49,360 --> 00:23:51,520 That must have been quite an intimidating thing. 412 00:23:51,520 --> 00:23:54,080 I do think they were intimidated, but they were also determined. 413 00:23:55,440 --> 00:24:00,800 That is Professor Penny Axelrad from the University of Colorado. 414 00:24:00,800 --> 00:24:04,800 She's going to help me understand how trying to track Sputnik 415 00:24:04,800 --> 00:24:07,600 unlocked global navigation. 416 00:24:07,600 --> 00:24:10,600 Presumably, people are doing their best to try and find it. 417 00:24:10,600 --> 00:24:12,160 How are they doing that? 418 00:24:12,160 --> 00:24:15,040 So most people are looking at it with optical telescopes, right, 419 00:24:15,040 --> 00:24:16,600 they're looking at it at night. 420 00:24:16,600 --> 00:24:20,320 But these two young researchers had the idea of listening 421 00:24:20,320 --> 00:24:21,880 to the radio signals. 422 00:24:23,480 --> 00:24:27,480 Those young researchers, George Weiffenbach and William Guier, 423 00:24:27,480 --> 00:24:30,960 listened to Sputnik's beeping radio signal. 424 00:24:30,960 --> 00:24:33,160 And they noticed that it was changing. 425 00:24:33,160 --> 00:24:34,840 And they could measure 426 00:24:34,840 --> 00:24:38,680 the signal coming from Sputnik and measure how its frequency 427 00:24:38,680 --> 00:24:40,200 was shifting as it went by. 428 00:24:40,200 --> 00:24:42,400 You know, the Doppler shift. 429 00:24:42,400 --> 00:24:45,920 Here's the Doppler shift, helpfully demonstrated by our cameraman 430 00:24:45,920 --> 00:24:47,800 standing a bit too close to the kerb. 431 00:24:47,800 --> 00:24:49,720 CAR HORN BEEPS 432 00:24:49,720 --> 00:24:52,160 PITCH LOWERS 433 00:24:52,160 --> 00:24:55,880 That change in pitch that you can hear - that's caused by the sound 434 00:24:55,880 --> 00:24:58,880 waves getting compressed as the car gets closer, which leads 435 00:24:58,880 --> 00:25:00,480 to a higher frequency. 436 00:25:01,640 --> 00:25:04,520 And then the sound waves getting stretched out 437 00:25:04,520 --> 00:25:07,840 as the car moves away, leading to a lower frequency. 438 00:25:07,840 --> 00:25:10,040 You hear this... 439 00:25:07,840 --> 00:25:10,040 IMITATES TONE CHANGING 440 00:25:10,040 --> 00:25:11,480 ..kind of sound. 441 00:25:11,480 --> 00:25:14,760 They could use that information to actually figure out the orbit 442 00:25:14,760 --> 00:25:17,400 and then predict it, when it would come up next again. 443 00:25:17,400 --> 00:25:20,720 And so Sputnik's doppler shift could tell them exactly 444 00:25:20,720 --> 00:25:22,920 where it was at any moment. 445 00:25:24,080 --> 00:25:26,960 But it was what happened next that would change the history 446 00:25:26,960 --> 00:25:28,600 of navigation forever. 447 00:25:29,800 --> 00:25:32,480 Well, the next really cool thing was this... 448 00:25:32,480 --> 00:25:34,200 ..this guy McClure... 449 00:25:34,200 --> 00:25:36,160 That's Frank McClure. 450 00:25:36,160 --> 00:25:39,200 He was the supervisor of Guier and Weiffenbach, 451 00:25:39,200 --> 00:25:41,560 and he was working with the US Navy 452 00:25:41,560 --> 00:25:44,600 to try and develop a new navigation system. 453 00:25:44,600 --> 00:25:47,520 And he thought maybe, just maybe, 454 00:25:47,520 --> 00:25:50,240 he could turn Sputnik's doppler shift 455 00:25:50,240 --> 00:25:52,760 to the US's advantage. 456 00:25:52,760 --> 00:25:56,400 If you knew the orbit of the transmitter, then you could use 457 00:25:56,400 --> 00:25:58,960 those same measurements to figure out where you are. 458 00:25:58,960 --> 00:26:01,440 OK. Right? Because you do it upside down. 459 00:26:01,440 --> 00:26:04,640 Yeah. So if you know what the motion is and you make those same 460 00:26:04,640 --> 00:26:07,440 measurements on the ground, it can tell you where you're located. 461 00:26:07,440 --> 00:26:10,600 That's quite a big conceptual leap, then, isn't it? Yeah, absolutely. 462 00:26:12,120 --> 00:26:15,560 The doppler shift means that the sound of Sputnik changes 463 00:26:15,560 --> 00:26:17,680 as it goes overhead. 464 00:26:17,680 --> 00:26:21,360 McClure realised that you can use that to work out that moment 465 00:26:21,360 --> 00:26:22,880 when it is right above you. 466 00:26:22,880 --> 00:26:25,840 And if you know where Sputnik is supposed to be in its orbit 467 00:26:25,840 --> 00:26:29,640 at that moment, well, then you can pinpoint your location on Earth. 468 00:26:30,880 --> 00:26:34,120 This was super-important because when you're out in the middle 469 00:26:34,120 --> 00:26:38,160 of the ocean... Like if you're in a nuclear submarine, say. 470 00:26:38,160 --> 00:26:41,520 ..there's very few landmarks, there's nothing to use. 471 00:26:41,520 --> 00:26:44,200 And so having the idea that you could use a satellite 472 00:26:44,200 --> 00:26:45,800 to provide navigation information 473 00:26:45,800 --> 00:26:48,880 anywhere on the Earth, right, was huge. Wow. 474 00:26:51,280 --> 00:26:55,720 And yet we were still a long way away from GPS as we know it today. 475 00:26:58,120 --> 00:27:01,640 McClure's prototype system could only pinpoint your position 476 00:27:01,640 --> 00:27:05,920 to within 900m, which is not ideal for finding a street address 477 00:27:05,920 --> 00:27:08,240 for a pizza delivery. 478 00:27:11,200 --> 00:27:14,680 But change was coming, thanks to a new-fangled 479 00:27:14,680 --> 00:27:16,920 tech called computing - 480 00:27:16,920 --> 00:27:19,520 and one woman from rural Virginia. 481 00:27:20,680 --> 00:27:23,400 You know, when I went to college and high school, 482 00:27:23,400 --> 00:27:27,040 the biggest thing we ever had was a calculator that sat 483 00:27:27,040 --> 00:27:28,240 on your desk. 484 00:27:28,240 --> 00:27:30,880 This is Dr Gladys West. 485 00:27:30,880 --> 00:27:33,080 She grew up working on her family's farm 486 00:27:33,080 --> 00:27:34,800 in the 1930s. 487 00:27:36,560 --> 00:27:38,560 And it was always hot. 488 00:27:38,560 --> 00:27:42,360 And then you feel like, "Well, I just can't take this any more! 489 00:27:43,880 --> 00:27:46,200 "I will be leaving here as soon 490 00:27:46,200 --> 00:27:49,440 "as I can get able to... know how to get out!" 491 00:27:50,800 --> 00:27:54,920 Gladys had a talent for mathematics, and her hard work paid off 492 00:27:54,920 --> 00:28:00,440 when, in 1956, she was hired at Dahlgren, a US naval base. 493 00:28:01,760 --> 00:28:06,600 She was just the fourth black employee to be hired on the base, 494 00:28:06,600 --> 00:28:11,000 at a time when racial segregation was still in place. 495 00:28:11,000 --> 00:28:17,280 And so to go into a place completely empty of black people, you know. 496 00:28:17,280 --> 00:28:22,640 So therefore, you got to know how to rise above it and not let them 497 00:28:22,640 --> 00:28:24,080 drag you down. 498 00:28:24,080 --> 00:28:27,520 Don't act like you're black - act like you're Gladys. 499 00:28:30,120 --> 00:28:34,080 Gladys was given access to some of the most powerful computers 500 00:28:34,080 --> 00:28:35,680 of the era. 501 00:28:35,680 --> 00:28:40,400 I came to Dahlgren, and the computer that was so huge. 502 00:28:40,400 --> 00:28:43,880 And that's why they were doing the satellites 503 00:28:43,880 --> 00:28:45,760 and space work, I guess. 504 00:28:48,120 --> 00:28:52,160 So here is the problem that the US Navy had. 505 00:28:52,160 --> 00:28:55,960 If you want to build a GPS system that can really accurately 506 00:28:55,960 --> 00:29:00,320 pinpoint locations on Earth, then you need to know how close 507 00:29:00,320 --> 00:29:03,080 all of your satellites are to the ground. 508 00:29:03,080 --> 00:29:05,400 And that sounds like it should be simple. 509 00:29:05,400 --> 00:29:09,080 Except, unfortunately, the Earth isn't a perfect sphere. 510 00:29:09,080 --> 00:29:13,880 The land is covered in hills and mountains. And the oceans, too - 511 00:29:13,880 --> 00:29:17,480 they're constantly rising and sinking because of the tides. 512 00:29:17,480 --> 00:29:20,920 And that royally messes up the accuracy 513 00:29:20,920 --> 00:29:23,240 of your would-be GPS satellites. 514 00:29:24,800 --> 00:29:28,880 What they needed was a faithful computer model of the Earth 515 00:29:28,880 --> 00:29:32,440 that included all of those bumps and bulges. 516 00:29:32,440 --> 00:29:34,720 And they asked Gladys to build it. 517 00:29:36,320 --> 00:29:40,720 You get excited when you have developed something, a model, 518 00:29:40,720 --> 00:29:44,920 and then you have to check it, check it out, make sure that it works. 519 00:29:44,920 --> 00:29:49,560 Gladys spent 12 years collecting data on tidal changes 520 00:29:49,560 --> 00:29:52,040 and the Earth's irregular shape 521 00:29:52,040 --> 00:29:55,400 and then wrote a series of computer algorithms 522 00:29:55,400 --> 00:29:58,160 to correct the inaccuracies that they created 523 00:29:58,160 --> 00:29:59,800 for satellite signals. 524 00:30:01,520 --> 00:30:06,800 Finally, in 1978, the first experimental version of today's 525 00:30:06,800 --> 00:30:08,920 GPS was launched. 526 00:30:08,920 --> 00:30:13,040 It's amazing, what we have done. It just, it's exciting. 527 00:30:13,040 --> 00:30:15,160 This just makes me feel happy. 528 00:30:17,120 --> 00:30:21,800 Thanks to Gladys's precise modelling of the Earth, GPS was now accurate 529 00:30:21,800 --> 00:30:23,480 to within a few metres. 530 00:30:25,480 --> 00:30:28,240 It quickly became an essential military tool... 531 00:30:29,480 --> 00:30:32,120 ..but was still off limits to the likes of you, 532 00:30:32,120 --> 00:30:34,720 me and the pizza delivery guy. 533 00:30:35,880 --> 00:30:37,880 Until 1983, 534 00:30:37,880 --> 00:30:40,760 when tragedy struck. 535 00:30:40,760 --> 00:30:44,400 America has blamed the Russians for shooting down a South Korean 536 00:30:44,400 --> 00:30:47,960 jumbo jet, which is missing with 269 people. 537 00:30:47,960 --> 00:30:49,760 They're all presumed dead. 538 00:30:49,760 --> 00:30:53,360 Without access to the military's GPS system, 539 00:30:53,360 --> 00:30:56,400 the passenger plane had made a navigation error that would 540 00:30:56,400 --> 00:30:58,040 lead to disaster. 541 00:30:59,320 --> 00:31:02,440 It had unknowingly drifted into Soviet airspace 542 00:31:02,440 --> 00:31:04,040 and was shot down. 543 00:31:05,480 --> 00:31:08,840 President Reagan vowed to take action. 544 00:31:08,840 --> 00:31:11,960 Our prayers tonight are with the victims and their families 545 00:31:11,960 --> 00:31:13,840 in their time of terrible grief. 546 00:31:13,840 --> 00:31:15,600 Our hearts go out to them. 547 00:31:15,600 --> 00:31:18,600 Our immediate challenge to this atrocity is to ensure 548 00:31:18,600 --> 00:31:20,440 that we make the skies safer. 549 00:31:20,440 --> 00:31:23,440 To try and avoid similar tragedies, 550 00:31:23,440 --> 00:31:27,400 he announced that GPS would be made available to all. 551 00:31:30,280 --> 00:31:35,840 Ten years, two more US Presidents and $10 billion later, 552 00:31:35,840 --> 00:31:40,520 we got the GPS system used by most sat navs across the world today. 553 00:31:42,000 --> 00:31:45,120 At the roundabout, turn right to join the main road. 554 00:31:45,120 --> 00:31:46,560 Thank you, dear. 555 00:31:46,560 --> 00:31:48,360 She can't hear you, mate. 556 00:31:48,360 --> 00:31:50,720 Here's the thing that I keep noticing about these stories 557 00:31:50,720 --> 00:31:54,760 over and over again is that it's not just that the innovation 558 00:31:54,760 --> 00:31:57,200 happens in isolation. 559 00:31:57,200 --> 00:32:01,360 It's about everything coming together at a moment in time. 560 00:32:01,360 --> 00:32:05,480 So, here, all of the stuff about doppler effect and triangles 561 00:32:05,480 --> 00:32:09,000 and timecodes, all of that had been known for years and years and years 562 00:32:09,000 --> 00:32:11,000 before this system was launched, 563 00:32:11,000 --> 00:32:16,120 but it took the Cold War, the terrifying fear of Sputnik, 564 00:32:16,120 --> 00:32:20,960 that incredible need for innovation, for everything to come together, 565 00:32:20,960 --> 00:32:23,280 and this new technology to be created. 566 00:32:25,280 --> 00:32:30,160 GPS is undoubtedly a little bit of everyday genius on which 567 00:32:30,160 --> 00:32:33,680 food delivery apps completely depend. 568 00:32:33,680 --> 00:32:35,600 But on its own, it's not enough 569 00:32:35,600 --> 00:32:39,200 for that pizza I ordered to actually turn up at my house. 570 00:32:42,840 --> 00:32:46,200 For that, we're going to need more algorithms. 571 00:32:48,320 --> 00:32:51,600 Issuing instructions from back at base, 572 00:32:51,600 --> 00:32:55,120 they are in constant communication with Deliveroo riders to match them 573 00:32:55,120 --> 00:32:56,720 up with orders. 574 00:32:56,720 --> 00:32:59,760 Kind of like Tinder for your takeaway. 575 00:32:59,760 --> 00:33:04,240 It's the algorithms behind the delivery rider's display. 576 00:33:08,240 --> 00:33:12,640 That amazing variety of food and the speed of delivery offered 577 00:33:12,640 --> 00:33:13,880 by the order screen - 578 00:33:13,880 --> 00:33:16,920 all of that only works if there is a system 579 00:33:16,920 --> 00:33:20,760 that can call on a vast pool of freelance delivery riders 580 00:33:20,760 --> 00:33:24,920 that can constantly adapt to a whole range of changing factors. 581 00:33:30,960 --> 00:33:33,840 Do you think about how many moving parts there are? 582 00:33:33,840 --> 00:33:36,960 It's not just that you're at home on a Friday night, on your sofa, 583 00:33:36,960 --> 00:33:39,400 deciding whether or not you're going to have a takeaway, 584 00:33:39,400 --> 00:33:42,600 whether you're going to order a burger or pizza or a curry. 585 00:33:42,600 --> 00:33:44,280 You've got that happening 586 00:33:44,280 --> 00:33:47,240 thousands upon thousands upon thousands of times 587 00:33:47,240 --> 00:33:49,960 simultaneously across an entire city. 588 00:33:49,960 --> 00:33:52,880 You've also got to think about the perspective of the riders. 589 00:33:52,880 --> 00:33:55,080 Do they want to go out to work or not that night? 590 00:33:55,080 --> 00:33:57,760 You've got the restaurants, you've got the weather, 591 00:33:57,760 --> 00:34:00,680 you've got the thousands and thousands of moving parts, 592 00:34:00,680 --> 00:34:04,640 and Deliveroo has to navigate all of them simultaneously to work 593 00:34:04,640 --> 00:34:06,400 as effectively as possible. 594 00:34:06,400 --> 00:34:08,760 And it sounds really difficult. 595 00:34:09,800 --> 00:34:14,720 Luckily, they've got a secret weapon up their sleeve called Frank. 596 00:34:17,200 --> 00:34:18,760 And before you ask, 597 00:34:18,760 --> 00:34:22,040 no, Frank isn't just one really quick guy on a bike. 598 00:34:24,400 --> 00:34:27,640 So first of all, Frank is not a person. 599 00:34:28,680 --> 00:34:30,480 See? Told you. 600 00:34:30,480 --> 00:34:32,680 That's Devesh Mishra, Deliveroo's 601 00:34:32,680 --> 00:34:34,520 chief technology officer. 602 00:34:35,600 --> 00:34:39,760 So Frank is our super-smart machine-learning algorithm 603 00:34:39,760 --> 00:34:43,280 that helps us to ensure each order is delivered 604 00:34:43,280 --> 00:34:45,080 as efficiently as possible. 605 00:34:45,080 --> 00:34:49,160 Their algorithms have given us order screens with almost unlimited 606 00:34:49,160 --> 00:34:52,640 choice, and the system that they've devised to deliver 607 00:34:52,640 --> 00:34:56,960 all of that choice also depends on algorithms. 608 00:34:56,960 --> 00:34:59,440 Can't you just work out who the closest rider is 609 00:34:59,440 --> 00:35:01,160 and just get them to take the order? 610 00:35:01,160 --> 00:35:04,680 If you think about this one order in isolation and you assume 611 00:35:04,680 --> 00:35:09,040 that the customer is very close to the restaurant, then it may make 612 00:35:09,040 --> 00:35:12,240 sense to just assign the closest rider to that order. 613 00:35:12,240 --> 00:35:16,680 But in reality, at any given point of time, we have thousands 614 00:35:16,680 --> 00:35:20,640 and thousands of orders, and the delivery time depends 615 00:35:20,640 --> 00:35:25,800 on a lot of factors, which could be the particular meal 616 00:35:25,800 --> 00:35:29,600 that is being ordered, which could be the vehicle type 617 00:35:29,600 --> 00:35:32,480 of the rider, which could be the road conditions. 618 00:35:32,480 --> 00:35:37,480 All of that complexity behind the scenes pops up as a job offer 619 00:35:37,480 --> 00:35:39,080 on the rider's display. 620 00:35:40,360 --> 00:35:44,560 So this is the rider system which allows each rider to receive 621 00:35:44,560 --> 00:35:46,640 offer of orders from Deliveroo, 622 00:35:46,640 --> 00:35:48,240 and then they can decide 623 00:35:48,240 --> 00:35:50,680 whether to accept or reject the order. 624 00:35:50,680 --> 00:35:53,920 And this is a heat map. So if you are a rider in this particular area, 625 00:35:53,920 --> 00:35:55,800 it will show you whether it is busy. 626 00:35:55,800 --> 00:35:58,520 And then based on that, you can decide whether you want to work 627 00:35:58,520 --> 00:36:01,480 in this area, whether you want to accept the orders or not. 628 00:36:02,600 --> 00:36:05,320 The heat map on the rider's display shows riders 629 00:36:05,320 --> 00:36:09,640 where they're most likely to pick up a job, so they might choose 630 00:36:09,640 --> 00:36:12,080 to travel there from a quieter area, 631 00:36:12,080 --> 00:36:16,440 adding to the pool of riders in those areas that need them the most. 632 00:36:18,280 --> 00:36:22,200 Frank will then offer incoming jobs to the rider that it calculates 633 00:36:22,200 --> 00:36:26,480 will be fastest for each delivery, depending on everything from traffic 634 00:36:26,480 --> 00:36:30,040 conditions to whether you're on a bike or a moped. 635 00:36:30,040 --> 00:36:33,680 That offer is theirs for 100 seconds. 636 00:36:33,680 --> 00:36:36,960 If they don't take it, it goes to the next best-placed rider. 637 00:36:36,960 --> 00:36:40,760 And then Frank takes all of this into account to offer 638 00:36:40,760 --> 00:36:43,600 the order to the right set of riders. 639 00:36:43,600 --> 00:36:47,560 Frank's algorithms can process millions of data points a second... 640 00:36:50,120 --> 00:36:56,160 ..but 130 years ago, Indian office workers pioneered a delivery 641 00:36:56,160 --> 00:37:00,680 data management system that was equally robust, just with much 642 00:37:00,680 --> 00:37:02,480 smaller numbers involved. 643 00:37:02,480 --> 00:37:06,400 OK, I want you to imagine that it's the 1890s in India. 644 00:37:06,400 --> 00:37:08,560 Mumbai is absolutely booming. 645 00:37:08,560 --> 00:37:12,320 Thousands and thousands of workers are travelling up to three hours 646 00:37:12,320 --> 00:37:15,960 to get into the centre of the city to work every single day. 647 00:37:15,960 --> 00:37:19,320 That is too far to just nip home for lunch. 648 00:37:22,120 --> 00:37:26,960 But then a visionary entrepreneur called Mahadeo Havaji Bachche saw 649 00:37:26,960 --> 00:37:30,640 an opportunity for a whole new way of getting home-cooked lunches 650 00:37:30,640 --> 00:37:33,440 to the hungry office workers. 651 00:37:33,440 --> 00:37:37,160 Why not just pay somebody to go and pick up the food from your house 652 00:37:37,160 --> 00:37:40,400 and then deliver it to your desk when it's lunchtime? 653 00:37:40,400 --> 00:37:41,440 Pretty smart. 654 00:37:42,680 --> 00:37:45,520 It was the birth of an extraordinary food delivery system known 655 00:37:45,520 --> 00:37:47,280 as dabbawalla. 656 00:37:49,240 --> 00:37:52,840 It's so efficient that they still use it in Mumbai. 657 00:37:52,840 --> 00:37:57,000 Every day, around midday, dabbawallas deliver home-cooked meals 658 00:37:57,000 --> 00:37:59,680 to 200,000 city workers. 659 00:38:01,720 --> 00:38:04,840 Now, I know you're expecting me to travel to Mumbai to see 660 00:38:04,840 --> 00:38:08,680 if these lunches get to the right people. 661 00:38:08,680 --> 00:38:11,760 Unluckily for me, our production budget didn't quite stretch to that, 662 00:38:11,760 --> 00:38:14,760 so there is a small-scale version here in London to show me. 663 00:38:19,720 --> 00:38:23,880 Oh, my gosh. This smells...amazing! 664 00:38:23,880 --> 00:38:25,080 So what have we got here? 665 00:38:25,080 --> 00:38:28,240 So we've got dhal, there's rice, 666 00:38:28,240 --> 00:38:30,240 bhel and paneer. 667 00:38:30,240 --> 00:38:33,840 In India, people think of food as way of showing love. 668 00:38:33,840 --> 00:38:36,360 And I think the dabba system is part of that, 669 00:38:36,360 --> 00:38:37,800 it's the showing of love. 670 00:38:38,960 --> 00:38:42,640 This is food historian Monisha Bharadwaj. 671 00:38:42,640 --> 00:38:46,960 She grew up in Mumbai and knows this delivery system inside out. 672 00:38:48,240 --> 00:38:50,320 So tell me, how does the system work? 673 00:38:50,320 --> 00:38:54,240 The meal is prepared in the morning, usually by the lady of the house, 674 00:38:54,240 --> 00:38:58,080 and then the dabbawalla. Dabbawalla literally means the person 675 00:38:58,080 --> 00:38:59,720 who carries the box. 676 00:38:59,720 --> 00:39:03,880 He picks up this freshly cooked meal and he cycles to the nearest train 677 00:39:03,880 --> 00:39:07,520 station and it's handed over to the next dabbawalla, 678 00:39:07,520 --> 00:39:09,320 who takes the train. 679 00:39:09,320 --> 00:39:12,280 And when they arrive at the destination station, they hand 680 00:39:12,280 --> 00:39:15,680 it to another person who then actually delivers it to the office 681 00:39:15,680 --> 00:39:18,360 and to the desk, even. It's kind of like a relay. 682 00:39:18,360 --> 00:39:19,840 It gets handed over from one person 683 00:39:19,840 --> 00:39:21,440 to the next person to the next person 684 00:39:21,440 --> 00:39:24,760 and then seamlessly arrives at your desk, 685 00:39:24,760 --> 00:39:26,360 still warm. 686 00:39:26,360 --> 00:39:29,680 That means when your husband is at work, he's going to be able 687 00:39:29,680 --> 00:39:32,040 to eat food that was cooked at home. Does it ever happen 688 00:39:32,040 --> 00:39:34,920 the other way around, where you've got the men at home who are doing 689 00:39:34,920 --> 00:39:37,840 the cooking and they send it off to their wives in the city? Not yet. 690 00:39:38,920 --> 00:39:40,760 Maybe one day. Maybe one day. 691 00:39:41,800 --> 00:39:43,880 I mean, it does sound quite complicated, though. 692 00:39:43,880 --> 00:39:45,080 Do they not get mixed up? 693 00:39:45,080 --> 00:39:48,920 They don't because they've got this really wonderful system 694 00:39:48,920 --> 00:39:50,520 of marking the boxes. 695 00:39:50,520 --> 00:39:52,800 Can we do one? Can you show me how it works? 696 00:39:52,800 --> 00:39:55,240 Let's make our own little coding system here. 697 00:39:55,240 --> 00:39:57,280 Let's imagine that my husband has been up cooking 698 00:39:57,280 --> 00:40:00,640 early this morning. Lovely! So the pick-up is my house. 699 00:40:00,640 --> 00:40:03,880 The pick-up is your house. Let's put an H there, for Hannah. 700 00:40:03,880 --> 00:40:07,760 Then let's change colour, where the box is going to. 701 00:40:07,760 --> 00:40:09,440 So the destination train station. 702 00:40:09,440 --> 00:40:11,360 Sometimes I go into the BBC offices. 703 00:40:11,360 --> 00:40:13,840 So let's say, let's imagine that that's Oxford Circus. 704 00:40:13,840 --> 00:40:15,920 Lovely. So that's one for Oxford Circus. 705 00:40:15,920 --> 00:40:19,600 And then we're going to put the building and the desk. 706 00:40:19,600 --> 00:40:20,920 Is it W1A? 707 00:40:20,920 --> 00:40:22,520 Yeah, that's the BBC building. 708 00:40:22,520 --> 00:40:24,840 Which floor are you? I'm on the second floor. 709 00:40:24,840 --> 00:40:26,400 Second. 710 00:40:26,400 --> 00:40:28,520 Desk number? I'm quite low down the pecking order! 711 00:40:28,520 --> 00:40:30,280 So let's say it's nine. 712 00:40:30,280 --> 00:40:34,280 That's your box, and you can expect a hot meal at 11am. 713 00:40:34,280 --> 00:40:36,280 Oh, I hope he's picked my favourite! 714 00:40:38,120 --> 00:40:39,920 Using this coding system, 715 00:40:39,920 --> 00:40:42,480 mistakes are vanishingly rare. 716 00:40:43,600 --> 00:40:47,160 And I think that's why people use the dabbawalla system so much, 717 00:40:47,160 --> 00:40:48,880 because it is seamless, 718 00:40:48,880 --> 00:40:51,040 because it is a no-fail system. 719 00:40:51,040 --> 00:40:54,360 But this is a system that was working way before any technology 720 00:40:54,360 --> 00:40:56,080 or barcodes or anything like that. 721 00:40:56,080 --> 00:40:58,760 True. And also far more reliable. 722 00:41:00,880 --> 00:41:03,880 It works brilliantly when the dabbawallas are going back 723 00:41:03,880 --> 00:41:06,720 to the same handful of locations every day, 724 00:41:06,720 --> 00:41:09,960 but there is a limit to how many data entries you can make 725 00:41:09,960 --> 00:41:12,520 when you're writing on the side of a tin. 726 00:41:13,800 --> 00:41:16,680 Luckily for us, there was that new-fangled invention 727 00:41:16,680 --> 00:41:18,320 from the 1950s. 728 00:41:18,320 --> 00:41:22,360 The computer. Remember? The one that gave us GPS? 729 00:41:22,360 --> 00:41:25,160 Its knack for handling masses of computer data 730 00:41:25,160 --> 00:41:27,160 was also about to revolutionise 731 00:41:27,160 --> 00:41:30,640 how we transport everything from people to pizzas. 732 00:41:32,280 --> 00:41:36,000 And that revolution began in a cafe in Amsterdam. 733 00:41:37,720 --> 00:41:41,320 Back in 1956, there was a 26-year-old Dutchman 734 00:41:41,320 --> 00:41:43,320 called Edsger Dijkstra, 735 00:41:43,320 --> 00:41:46,400 and he'd gone shopping with his fiancee, decided to take 736 00:41:46,400 --> 00:41:50,000 a bit of a breather, stop for coffee, and for 20 minutes, 737 00:41:50,000 --> 00:41:53,400 rather than having a conversation with his best beloved, 738 00:41:53,400 --> 00:41:57,000 he decided instead to set himself a little mathematical puzzle, 739 00:41:57,000 --> 00:41:59,480 which is what you have to expect if you're going to date 740 00:41:59,480 --> 00:42:00,800 a mathematician. 741 00:42:00,800 --> 00:42:04,960 He decided to try and work out how to get from Rotterdam 742 00:42:04,960 --> 00:42:10,040 down here, all the way up to Groningen in the north-east, 743 00:42:10,040 --> 00:42:13,320 in the shortest path possible. 744 00:42:13,320 --> 00:42:16,560 The thing that makes this a tricky problem is that if you're going 745 00:42:16,560 --> 00:42:19,800 to get from here to there, there are lots of different potential routes 746 00:42:19,800 --> 00:42:21,080 that you could go through. 747 00:42:21,080 --> 00:42:25,520 And so finding the quickest one isn't necessarily that simple. 748 00:42:25,520 --> 00:42:31,520 So what Dijkstra did is work out this algorithm in his head. 749 00:42:33,280 --> 00:42:36,600 Dijkstra thought, if a computer knew all of the distances 750 00:42:36,600 --> 00:42:40,280 between neighbouring stations, well, then, by working through 751 00:42:40,280 --> 00:42:44,120 all possible combinations, it should be able to find the shortest route 752 00:42:44,120 --> 00:42:46,320 between any two stations. 753 00:42:48,080 --> 00:42:51,720 To show you what I mean, instead of using a 1950s computer, 754 00:42:51,720 --> 00:42:55,040 I'm going to use an equally cool invention from that era. 755 00:42:55,040 --> 00:42:56,600 Lego. 756 00:42:58,800 --> 00:43:03,120 This is my Lego map of the Dutch railway system. 757 00:43:03,120 --> 00:43:05,400 It's very flat, as you might expect. 758 00:43:05,400 --> 00:43:09,280 We also have all of the cities marked out as the yellow lights 759 00:43:09,280 --> 00:43:10,960 that you can see here, 760 00:43:10,960 --> 00:43:16,120 and the white dots are the train tracks that run between each 761 00:43:16,120 --> 00:43:17,720 of the locations. 762 00:43:17,720 --> 00:43:20,920 So, if we start with a really simple example, let's say that you wanted 763 00:43:20,920 --> 00:43:24,040 to go between Utrecht and Zwolle. 764 00:43:24,040 --> 00:43:27,200 Now, you don't need a computer to solve this one for you, 765 00:43:27,200 --> 00:43:30,160 because even a child could tell you that the shortest path 766 00:43:30,160 --> 00:43:32,320 is just going to be a straight line between the two. 767 00:43:32,320 --> 00:43:34,320 Like this right here. 768 00:43:34,320 --> 00:43:37,240 But where Dijkstra's algorithm really comes into its own 769 00:43:37,240 --> 00:43:39,760 is when you're looking at a much more complicated example. 770 00:43:39,760 --> 00:43:44,160 So, let's imagine, as Dijkstra himself did, that you are trying 771 00:43:44,160 --> 00:43:46,840 to get from Rotterdam over here, 772 00:43:46,840 --> 00:43:50,040 all the way over to Groningen over here. 773 00:43:50,040 --> 00:43:51,480 Suddenly a lot trickier. 774 00:43:52,640 --> 00:43:56,360 So what he did was feed all the distances between neighbouring 775 00:43:56,360 --> 00:44:01,400 stations into a computer, and got it to start adding up the segments. 776 00:44:01,400 --> 00:44:05,600 Now, the algorithm has to search through every possible combination 777 00:44:05,600 --> 00:44:11,720 of stops and add up the distances to get from the origin all the way 778 00:44:11,720 --> 00:44:13,240 to the destination, 779 00:44:13,240 --> 00:44:16,800 while keeping hold in its memory of the shortest 780 00:44:16,800 --> 00:44:18,600 one that it's found so far. 781 00:44:18,600 --> 00:44:22,160 And when it's exhausted all of the possibilities, it knows 782 00:44:22,160 --> 00:44:27,400 for definite that it has the shortest possible path, 783 00:44:27,400 --> 00:44:30,200 namely, in this particular example, 784 00:44:30,200 --> 00:44:33,640 the one that runs through here like this. 785 00:44:34,880 --> 00:44:37,960 When Dijkstra went public with his computer algorithm, 786 00:44:37,960 --> 00:44:41,640 its universal usefulness was immediately obvious. 787 00:44:43,000 --> 00:44:46,360 What Dykstra had done is he'd created something 788 00:44:46,360 --> 00:44:50,520 that could be understood and automated by a computer, 789 00:44:50,520 --> 00:44:54,720 and the core of his idea is so powerful 790 00:44:54,720 --> 00:44:59,080 because it can be applied to any rail network or any road network 791 00:44:59,080 --> 00:45:00,720 in the entire world. 792 00:45:00,720 --> 00:45:05,200 His algorithm is still at the heart of modern-day GPS, 793 00:45:05,200 --> 00:45:07,040 of Google Maps 794 00:45:07,040 --> 00:45:09,480 and even the Deliveroo app. 795 00:45:11,560 --> 00:45:15,360 It's a key part of Frank, that computer brain coordinating 796 00:45:15,360 --> 00:45:18,320 the display used by every Deliveroo rider. 797 00:45:19,480 --> 00:45:21,720 But Frank does a lot more than just remember 798 00:45:21,720 --> 00:45:24,160 the distances between points. 799 00:45:24,160 --> 00:45:28,880 It factors in tonnes of other data that affects delivery time. 800 00:45:30,240 --> 00:45:33,400 So we have a model which also predicts how much time 801 00:45:33,400 --> 00:45:35,160 rider has to wait for that order. 802 00:45:35,160 --> 00:45:40,360 So, for example, if the item only takes three minutes to cook, 803 00:45:40,360 --> 00:45:43,360 you don't want to assign it to a rider who is just two minutes 804 00:45:43,360 --> 00:45:46,200 away from the restaurant, because then they will be waiting. 805 00:45:46,200 --> 00:45:47,600 Right? If you're the rider, 806 00:45:47,600 --> 00:45:50,160 you want to be waiting as little time as possible. 807 00:45:50,160 --> 00:45:53,280 If you're the restaurant, you want to get as many orders in 808 00:45:53,280 --> 00:45:56,920 as possible. And so it sounds like you're having to work to balance 809 00:45:56,920 --> 00:45:59,800 all those different priorities. That's exactly right. 810 00:46:01,640 --> 00:46:06,000 This thing has got layers and layers of little optimizations 811 00:46:06,000 --> 00:46:08,520 that need to all work seamlessly together. 812 00:46:08,520 --> 00:46:12,800 Now, I know I am not allowed to give you my full lecture 813 00:46:12,800 --> 00:46:17,160 on stochastic machine learning models, but believe me when I tell 814 00:46:17,160 --> 00:46:22,040 you that there is an awful lot of maths going on in this building. 815 00:46:24,440 --> 00:46:26,880 Frank and the Deliveroo rider's display 816 00:46:26,880 --> 00:46:28,720 are an extraordinary bit of tech. 817 00:46:30,240 --> 00:46:33,720 But there is one more part of the delivery chain that's missing. 818 00:46:35,520 --> 00:46:39,360 We've arrived at the fourth clever tech that makes the food 819 00:46:39,360 --> 00:46:41,160 delivery app work. 820 00:46:41,160 --> 00:46:44,840 It's more lo-fi than the others, but still crucial. 821 00:46:46,040 --> 00:46:49,080 Because what would be the point in all of this amazing tech 822 00:46:49,080 --> 00:46:51,760 if your dinner turned up cold? 823 00:46:51,760 --> 00:46:54,680 Introducing the humble thermal bag. 824 00:47:00,160 --> 00:47:03,000 So, your takeaway's ready to go... 825 00:47:03,000 --> 00:47:04,320 All right. 826 00:47:04,320 --> 00:47:07,520 ..and the best rider for the job - in this case, that's me - 827 00:47:07,520 --> 00:47:09,200 has picked it up... 828 00:47:09,200 --> 00:47:10,480 Thank you. 829 00:47:10,480 --> 00:47:14,000 Just, if you could, like, stay upright, it would be great. 830 00:47:14,000 --> 00:47:15,840 Are you ready? Shall we go? 831 00:47:15,840 --> 00:47:17,120 Yeah. 832 00:47:17,120 --> 00:47:20,120 ..and accompanied by my fellow rider, Atlanta. 833 00:47:20,120 --> 00:47:23,600 The only thing that can save your evening from a cold, soggy 834 00:47:23,600 --> 00:47:26,200 takeaway is the thermal bag it's in. 835 00:47:28,440 --> 00:47:29,800 Nice one. 836 00:47:30,960 --> 00:47:32,120 Hi! 837 00:47:32,120 --> 00:47:33,920 There you go. 838 00:47:33,920 --> 00:47:36,280 Enjoy! Thank you. Thanks. 839 00:47:36,280 --> 00:47:37,840 Didn't give me a tip. 840 00:47:39,000 --> 00:47:41,040 What's that about? Only a few do. 841 00:47:42,200 --> 00:47:45,840 So, Hannah, I hear you ask, how does the thermal bag work? 842 00:47:45,840 --> 00:47:47,600 Great question. 843 00:47:48,960 --> 00:47:51,800 My director handed me a script and it's got loads of pages 844 00:47:51,800 --> 00:47:56,840 on it about how clever it is and all of the physics and blah, blah, blah. 845 00:47:56,840 --> 00:47:59,800 But then he said, "If you prefer, 846 00:47:59,800 --> 00:48:01,520 "I can give you a knife, 847 00:48:01,520 --> 00:48:04,240 "and you can just slash it open and see what's inside." 848 00:48:04,240 --> 00:48:06,120 Ooh! Sharp knife. 849 00:48:06,120 --> 00:48:07,360 OK. 850 00:48:08,560 --> 00:48:11,040 I enjoyed that maybe a little bit too much. 851 00:48:11,040 --> 00:48:13,160 OK, so now you can see these layers. 852 00:48:13,160 --> 00:48:16,760 So you've got this very thin layer of foam in the middle. 853 00:48:16,760 --> 00:48:20,360 Now, this foam has quite an important job because, of course, 854 00:48:20,360 --> 00:48:23,960 if you just put your takeaway in a metal box, 855 00:48:23,960 --> 00:48:26,040 metal, as a solid, 856 00:48:26,040 --> 00:48:29,480 and particularly a metallic one, would just steal away 857 00:48:29,480 --> 00:48:31,760 all of the heat from your food really quickly. 858 00:48:31,760 --> 00:48:36,480 On the other end of the spectrum, air is a really good insulator. 859 00:48:36,480 --> 00:48:41,800 So this foam has lots of little air pockets in it to ensure 860 00:48:41,800 --> 00:48:46,280 that your food is wrapped in this nice air bubble blanket. 861 00:48:47,760 --> 00:48:51,280 This silver actually plays a very special role 862 00:48:51,280 --> 00:48:54,880 because if you just had a black inside to the bag, 863 00:48:54,880 --> 00:48:58,040 black is very good at sucking away thermal energy, 864 00:48:58,040 --> 00:48:59,720 which means that your food 865 00:48:59,720 --> 00:49:01,840 would be cold once it arrived at your door. 866 00:49:01,840 --> 00:49:06,120 But the silver reflects it so that your food stays as hot 867 00:49:06,120 --> 00:49:08,800 as possible for as long as possible. 868 00:49:10,160 --> 00:49:14,160 So, on the outside it is covered in a breathable plastic, 869 00:49:14,160 --> 00:49:17,880 and it means that your food will stay waterproof 870 00:49:17,880 --> 00:49:19,760 if it's raining, on the outside. 871 00:49:19,760 --> 00:49:22,520 But this is also breathable, 872 00:49:22,520 --> 00:49:24,520 which means that any of the steam 873 00:49:24,520 --> 00:49:28,760 that's inside with your delicious takeaway is able to escape 874 00:49:28,760 --> 00:49:32,720 so that you end up with warm and not soggy food. 875 00:49:33,760 --> 00:49:34,920 It's pretty neat. 876 00:49:36,520 --> 00:49:39,800 And this unassuming design marvel was born in the decade 877 00:49:39,800 --> 00:49:42,600 that gave us Knight Rider and the A-Team. 878 00:49:44,840 --> 00:49:46,400 The 1980s. 879 00:49:49,520 --> 00:49:51,920 I'm an American, so I love pizza. 880 00:49:54,600 --> 00:49:56,480 That is Ingrid Kosar. 881 00:49:57,480 --> 00:50:01,560 She invented the first mass-produced thermal bag for pizza delivery. 882 00:50:02,680 --> 00:50:07,360 The state of pizza delivery in the '80s was pretty different. 883 00:50:07,360 --> 00:50:11,160 You know, they used a lot of blankets just to wrap 884 00:50:11,160 --> 00:50:13,360 around the pizzas to keep them warm. 885 00:50:13,360 --> 00:50:16,680 Everybody in the pizza business is looking for a way to deliver 886 00:50:16,680 --> 00:50:18,640 oven-fresh pizza! 887 00:50:18,640 --> 00:50:21,840 Some restaurants went to extremes. 888 00:50:21,840 --> 00:50:24,600 There was a company that had an electric heater 889 00:50:24,600 --> 00:50:27,080 that they would dry out the pizzas. 890 00:50:27,080 --> 00:50:30,440 They'd put that on the backseat of the car, but 891 00:50:30,440 --> 00:50:32,080 it would catch on fire a lot. 892 00:50:33,440 --> 00:50:38,680 So there was a need. There was a need for a delivery bag. 893 00:50:40,600 --> 00:50:44,920 So, while working her day job in a construction company, Ingrid 894 00:50:44,920 --> 00:50:48,680 thought outside of the pizza box and began to research 895 00:50:48,680 --> 00:50:50,000 high-tech fabrics. 896 00:50:52,000 --> 00:50:57,000 One of these called Thinsulate was just taking off in ski clothing. 897 00:50:58,200 --> 00:51:02,440 And if it could keep skiers warm, maybe it would be ideal for pizzas. 898 00:51:03,840 --> 00:51:05,520 So she got sewing. 899 00:51:07,360 --> 00:51:09,960 I had taught myself to sew years ago. 900 00:51:09,960 --> 00:51:13,840 Every day I'd go through a box of needles. It was pretty tough. 901 00:51:13,840 --> 00:51:19,000 The pizza pouch is the first bag I made. 902 00:51:19,000 --> 00:51:21,600 This one holds two 16-inch pizzas. 903 00:51:22,680 --> 00:51:25,840 This was made for Papa Johns. 904 00:51:25,840 --> 00:51:29,040 They were thinking of delivering beer with their pizzas. 905 00:51:29,040 --> 00:51:30,880 It was just a sample bag. 906 00:51:30,880 --> 00:51:34,960 And we made it to fit two six-packs of beer, which is really nice 907 00:51:34,960 --> 00:51:39,280 to have on hand, just in case you're going to a party. 908 00:51:41,120 --> 00:51:44,680 Ingrid approached Domino's Pizza chain with her new designs, 909 00:51:44,680 --> 00:51:47,000 and they set her a test. 910 00:51:47,000 --> 00:51:48,840 To pass the test, 911 00:51:48,840 --> 00:51:53,920 Domino's told us the pizza had to be at 140 degrees or higher 912 00:51:53,920 --> 00:51:56,480 after 30 minutes in the bag. 913 00:51:58,080 --> 00:52:02,240 Ingrid started testing her prototypes using a thermometer, 914 00:52:02,240 --> 00:52:06,120 a timer, and, well, far more pizza than she could possibly eat 915 00:52:06,120 --> 00:52:07,800 on her own. 916 00:52:07,800 --> 00:52:09,800 All my neighbours were really happy. 917 00:52:09,800 --> 00:52:11,440 Yeah, they loved me. 918 00:52:11,440 --> 00:52:13,320 BEEPER 919 00:52:13,320 --> 00:52:15,160 And then, in 1983... 920 00:52:15,160 --> 00:52:16,960 We passed the test. 921 00:52:16,960 --> 00:52:19,400 Ingrid's tests succeeded. 922 00:52:19,400 --> 00:52:25,040 She set up a business and filed a patent and, a year later, 923 00:52:25,040 --> 00:52:27,280 persuaded Domino's to take her on. 924 00:52:28,760 --> 00:52:29,920 Oh, my God. 925 00:52:29,920 --> 00:52:34,640 When you get your first good-sized order, there's no feeling like it. 926 00:52:34,640 --> 00:52:37,640 You're just... The joy 927 00:52:37,640 --> 00:52:39,920 just, like, jumps through your body. 928 00:52:39,920 --> 00:52:42,800 And it's...it's a high. 929 00:52:42,800 --> 00:52:45,040 It's a real high. I don't need a drug, 930 00:52:45,040 --> 00:52:46,440 I just need an order. 931 00:52:47,560 --> 00:52:50,800 Ingrid's basic design has stayed with us to this day. 932 00:52:53,680 --> 00:52:57,200 There has been an explosion of these bags in the last ten years 933 00:52:57,200 --> 00:53:00,560 as the food delivery market has grown bigger than ever. 934 00:53:02,120 --> 00:53:05,320 And so, thanks to this plucky thermal bag, 935 00:53:05,320 --> 00:53:08,920 takeaways across the land are successfully reaching 936 00:53:08,920 --> 00:53:11,600 their destinations hot and tasty. 937 00:53:14,880 --> 00:53:17,720 Where technology goes, food follows. 938 00:53:20,520 --> 00:53:24,080 And food delivery might just be about to get even more sci-fi... 939 00:53:25,960 --> 00:53:29,800 ..as tech starts to take people out of the equation altogether. 940 00:53:31,800 --> 00:53:34,880 Because the future is automation. 941 00:53:42,920 --> 00:53:47,520 I'm meeting a burger-making bot that has already proven itself 942 00:53:47,520 --> 00:53:52,440 in the cut-throat world of the San Francisco food scene. 943 00:53:54,800 --> 00:53:57,680 I think ultimately flipping a burger over and over every day isn't 944 00:53:57,680 --> 00:54:00,920 the culmination of the human spirit. I can tell you that from experience. 945 00:54:00,920 --> 00:54:04,200 I grew up flipping burgers. I think I've made like 30,000 in my life. 946 00:54:06,440 --> 00:54:10,320 That is Alex Vardakostas, CEO of Creator Burger. 947 00:54:12,120 --> 00:54:14,480 And this is his creation. 948 00:54:15,720 --> 00:54:20,280 This burger-making robot can create a customised burger to your spec 949 00:54:20,280 --> 00:54:23,800 to eat here or be delivered direct to your door. 950 00:54:26,080 --> 00:54:28,720 I have to say, there's a little bit of me that's disappointed 951 00:54:28,720 --> 00:54:30,360 that it's not like a humanoid robot, 952 00:54:30,360 --> 00:54:32,200 like flipping burgers like this. 953 00:54:32,200 --> 00:54:34,920 There are different ways of making robots. All right, 954 00:54:34,920 --> 00:54:36,360 so how does it work, then? 955 00:54:36,360 --> 00:54:40,240 OK, so this thing actually takes chunks of steak 956 00:54:40,240 --> 00:54:41,720 and grinds it to order for you. 957 00:54:41,720 --> 00:54:44,720 Then it uses a whole bunch of sensors and algorithms to cook it. 958 00:54:44,720 --> 00:54:48,040 Our robot cooks the patty at the top and bottom at the same time. 959 00:54:48,040 --> 00:54:51,240 So it saves time. Cooks it about 40% faster. 960 00:54:51,240 --> 00:54:53,240 It also lets us use the smash technique, 961 00:54:53,240 --> 00:54:56,360 which allows you to get a better what's called a Maillard reaction, 962 00:54:56,360 --> 00:54:58,880 a nice crust on the outside. That's the browning, isn't it? 963 00:54:58,880 --> 00:55:01,680 Yeah, it's delicious. OK, you have thought about this. 964 00:55:01,680 --> 00:55:04,400 What are the big advantages of it for you, as the owner? 965 00:55:04,400 --> 00:55:07,880 I would say, with automation typically, you get savings 966 00:55:07,880 --> 00:55:11,560 and efficiencies, but you also get advantages on the quality side, too, 967 00:55:11,560 --> 00:55:14,160 and so that's why automation can be a win-win. 968 00:55:14,160 --> 00:55:16,880 You've described this as essentially the most delicious burger 969 00:55:16,880 --> 00:55:19,320 of all-time. You have to see for yourself. Indeed, I will. 970 00:55:20,440 --> 00:55:22,000 It's time. Moment of truth. 971 00:55:25,040 --> 00:55:26,840 MUFFLED: It's really good! 972 00:55:27,920 --> 00:55:29,440 I need a napkin. 973 00:55:29,440 --> 00:55:31,000 That was a juicy burger. 974 00:55:31,000 --> 00:55:32,480 Let me take one more bite, 975 00:55:32,480 --> 00:55:34,360 because that was a bit of a messy bite. 976 00:55:35,880 --> 00:55:38,360 Actually, when you taste the meat, 977 00:55:38,360 --> 00:55:40,400 I mean, you barely have to chew. 978 00:55:40,400 --> 00:55:43,000 Just falls apart. It does. It just pulls apart. 979 00:55:43,000 --> 00:55:44,440 That's really nice. 980 00:55:44,440 --> 00:55:46,200 I hate to say it, but the machine makes 981 00:55:46,200 --> 00:55:48,200 a much better burger than I do. Oh, really? 982 00:55:48,200 --> 00:55:50,640 Does it make a better burger than ALL humans do? 983 00:55:50,640 --> 00:55:54,800 No, absolutely not. But we can make 120 of them an hour. 984 00:55:55,960 --> 00:55:58,920 Alex needs fewer staff and less kitchen space, 985 00:55:58,920 --> 00:56:03,080 and that brings his gourmet burgers down to around $7 each, 986 00:56:03,080 --> 00:56:06,160 in line with the big burger chains. 987 00:56:07,600 --> 00:56:10,200 Automation, while it makes a very good burger, 988 00:56:10,200 --> 00:56:12,880 it's not universally positive, though, right? 989 00:56:12,880 --> 00:56:15,400 We do employ less people than a normal restaurant. 990 00:56:15,400 --> 00:56:18,320 And so I think if people are saying, "Hey, we need to put people back 991 00:56:18,320 --> 00:56:21,040 "behind a hot griddle," and reduce the quality of food 992 00:56:21,040 --> 00:56:23,960 in doing so, I think we're having the wrong conversation. 993 00:56:23,960 --> 00:56:26,600 If you think about it, we're not going to have a future with 994 00:56:26,600 --> 00:56:29,200 like flying cars and a cure for cancer and some poor guy's 995 00:56:29,200 --> 00:56:31,800 still standing in front of a griddle like, "Still at it!" 996 00:56:31,800 --> 00:56:34,280 For us, I think the important thing is to ride that wave 997 00:56:34,280 --> 00:56:37,520 of inevitability that makes for a better food product for everybody. 998 00:56:37,520 --> 00:56:40,760 I have to say, the science that's gone into that machine 999 00:56:40,760 --> 00:56:43,680 is genuinely impressive. 1000 00:56:43,680 --> 00:56:46,040 But when it comes to the conversations 1001 00:56:46,040 --> 00:56:48,200 about the benefits of automation, 1002 00:56:48,200 --> 00:56:50,200 and I think that actually automation 1003 00:56:50,200 --> 00:56:52,640 really can destroy entire communities of people 1004 00:56:52,640 --> 00:56:54,520 when it's not handled carefully, 1005 00:56:54,520 --> 00:56:57,520 but I also think that we live in a world now where, if something 1006 00:56:57,520 --> 00:57:00,840 is technologically possible and it's more profitable 1007 00:57:00,840 --> 00:57:03,120 and increases quality, then... 1008 00:57:05,280 --> 00:57:06,680 ..it's kind of inevitable. 1009 00:57:07,880 --> 00:57:10,680 It WAS quite a tasty burger. Can't deny it. 1010 00:57:12,680 --> 00:57:16,480 It's clear that food delivery is here to stay, but the technology 1011 00:57:16,480 --> 00:57:19,480 to make it happen is forever moving on, 1012 00:57:19,480 --> 00:57:22,080 from the dabbawallas of Mumbai 1013 00:57:22,080 --> 00:57:24,840 to the arrival of the telephone, 1014 00:57:24,840 --> 00:57:27,480 the internet, and the extraordinary 1015 00:57:27,480 --> 00:57:30,000 sophistication of the algorithms 1016 00:57:30,000 --> 00:57:32,040 behind the most modern apps. 1017 00:57:33,320 --> 00:57:36,720 You know, when it came to filming this particular episode, 1018 00:57:36,720 --> 00:57:38,480 it was quite a challenge at times. 1019 00:57:38,480 --> 00:57:40,320 Unlike all of the other objects 1020 00:57:40,320 --> 00:57:42,520 that we are covering in this series, 1021 00:57:42,520 --> 00:57:45,560 there's not something that I can...I can get into 1022 00:57:45,560 --> 00:57:47,320 and physically show you. 1023 00:57:47,320 --> 00:57:49,560 It's a different kind of innovation. 1024 00:57:49,560 --> 00:57:53,280 It's an innovation that really belongs in the modern era 1025 00:57:53,280 --> 00:57:58,120 of technology, one that has changed the way we work, 1026 00:57:58,120 --> 00:58:00,600 but also the way that we eat. 1027 00:58:01,920 --> 00:58:04,240 Next time, the virtual assistant... 1028 00:58:04,240 --> 00:58:08,240 Alexa, do you mind if I use a screwdriver to open you? 1029 00:58:08,240 --> 00:58:10,440 Sorry, I'm not sure about that. 1030 00:58:10,440 --> 00:58:14,400 ..is the story of secret tech used to hunt U-boats. 1031 00:58:14,400 --> 00:58:18,480 A 1940s starlet who pioneered the science behind Wi-Fi... 1032 00:58:18,480 --> 00:58:21,240 I could promise you so many things. 1033 00:58:21,240 --> 00:58:25,440 ..oh, and a 1970s sci-fi villain who inspired today's virtual voices. 1034 00:58:25,440 --> 00:58:29,600 I'm sorry, Dave. I'm afraid I can't do that. 1035 00:58:30,960 --> 00:58:33,720 To explore the interesting dilemmas behind some familiar 1036 00:58:33,720 --> 00:58:35,160 pieces of technology, and hear 1037 00:58:35,160 --> 00:58:37,600 some of my own thoughts, go to... 1038 00:58:40,120 --> 00:58:42,960 ..and follow the links to the Open University.