Silvija Seres: Hello, and welcome to Lørn.Tech. My name is Silvija Seres, and our topic today is Internet of Things. My guest is Paul Houghton from a company called Future Is in Finland. He came all the way to Oslo, to teach us about Internet of Things, and how to play with technology for innovation. 

Paul Houghton: Yes, and to enjoy the nice weather. 

Silvija: Enjoy the nice, rainy, dark weather in Oslo. Although you probably are in the same boat in Helsinki. Welcome, Paul. 

Paul: Thank you. 

Silvija: Maybe we should start with telling our customers about who you are and what you do. 

Paul: I’m an engineer working in a consulting company that does a lot of general purpose mobile and web applications, as well as design. We are also helping with business change. My particular area is emerging technologies in Internet of Things. Let’s say everything else after you leave the square rectangular screen that you are used to using for digital services. The physical world. 

Silvija: People who know communication tell me that we are too narrow minded in the way we think of computers and human interaction. We are thinking keyboards and screens, but now we are moving into, basically, both voice and the physical world where we interact with things. Is that what you are talking about? 

Paul: Yes, I am thinking that there are a lot of possibilities, and there is some evidence for that we don’t necessarily have to tap on our phone in order to do something. You might have a dedicated magic wand with a few buttons that serves a need that you commonly have, or you could have a self driving car. These are all examples of where there is some intelligence inside a box and it is doing something for you, especially something with you, that perhaps allows you to do things that you couldn’t otherwise do in the real world, in the here and now. 

Silvija: So basically Future Is does not create a product, but they help people explore how to start implementing these products. These new ways of creating computers and digital stuff in their work. 

Paul: Yes. We have a variety of expertise, and mine is on the sidelines. It is not our main business.  But we are always trying to have no own products representing any given technology, but rather sitting on the same side of the table and being a useful advisor and hands on getting it done, once we have decided what should be done in digital services. 

Silvija: So when I talked to you initially, you introduced yourself as an American electrical engineer. I don’t know if an American electrical engineer is any different from others? You can tell us about that. 

Paul: Perhaps equally stubborn. 

Silvija: Equally stubborn - I like that! And serial entrepreneur? 

Paul: Yes. I have started a variety of companies for about ten years. And I have worked with business, but I found in the end, that I didn’t have any big breakthroughs. I wasn’t actually that motivated by money, and would rather work on interesting problems and helping people. 

Silvija: You like learning and getting others to learn. And we like that in Learn! 

Paul: Yeah, I like new things. 

Silvija: And then you said that about six or seven years ago, you were suddenly bored of mobiles. 

Paul: Yes. 

Silvija: And now you do embedding of machine learning into Internet of Things-products. So I’d like you to explain two things here. What is Internet of Things and how do you embed machine learning into that?

Paul: Internet of Things is what has been around for a long time. You have very small, inexpensive computers, that typically were in a factory or something like that. But they have become a lot less expensive and easier to put into a variety of objects like a home vacuum-cleaner that can wander around. And in addition to having a sort of tiny computer, usually there is also a radio, so it is a connected service. It’s not something that sits by itself anymore. And that can cost a few tens of kroner to add to any given device in terms of hardware cost. The real challenge is how to design that, so that it helps people and how to embed intelligence. Are you using traditional programming logic, or are you using these new algorithmes where you provide a lot of data and it finds out what the pattern is and how it can respond. For example if there is a camera, it can recognize cats. You can’t easily describe how it knows that it’s a cat, but you look at it and you know that it’s a cat. Those are good examples where a pattern with a lot of data can be extracted by modern algorithms which we did not have a few years ago. So that’s a really nice opportunity. 

Silvija: I have to do a very short digration. One of my favorite hackers in this country is a guy who is not necessarily a programmer by trade, but he was a guest on our drone-session and he works in our national broadcasting corporation. He had an introduction relatively recently about his smart home. He has made loads of plastic boxes with cheap chips that he bought from China, with these old Nokia phone-screens and connected them. Then he created cameras that take pictures of the bird-feeder, or that make his lights go blue when a satellite is passing over his house and he has made it so much fun. You realize that this is something that is not completely invisible and seamless. It is just people who are daring to play with this super cheap chip-technology and radio-technology, and trying to figure out how to get it to read our physical world. But can it be useful? I mean, pictures of birds and cats, and do people really need their fridge to order food, rather than having to remember that stuff themselves? 

Paul: I don’t know if need is the right question. No, they definitely don’t need it, but they may want it. I have a good example - I use the term intelligence augmentation, so we are not thinking about artificial intelligence, like there is a car that is going to drive, take my job and that it’s a robot killing the world. Rather that it should be something that gives you superpowers. It helps you and works intimately with you. Google Search is a good example. They studied the same field of cybernetics back in school, and it allows you to find things that you couldn’t otherwise find. Or Amazon Echo sitting in your home is increasingly popular. They use machine learning to recognize voice, so you can talk to appliances if that’s more convenient, or they can see, and they can perhaps help you drive if it’s particularly snowy, because they might be able to see things that you can’t. So that combination of a human trying to do something, and their extra scenes, it’s like people and dogs. We have evolved for years. It’s a better example. It’s nice to have a light turn blue when a satellite flies over. It’s fun, and I love playing with maker-stuff. I have a room with three 3D-printers, boxes of electronics and I 3D-print legos, so that I can embed them into my early prototypes. Things which you can’t buy in a store. This is fun, but it’s not where people are going to get value. It’s entertainment for nerds. 

Silvija: But then maybe that’s how nerds learn, and that’s how the rest of the world needs to learn. 

Paul: Part of our job is to find out where it’s useful and where it isn’t. I’m at the moment looking at how to use voice to text, so that if you put a box in the middle of the meeting room and you record everything that’s happened, who said what, and with their permission you can turn that into text and make it rapidly searchable and create new ways to find experts inside of your company. If that makes your life a little bit easier, then that’s a good thing. 

Silvija: Absolutely. I am actually surprised that we don’t have that. 

Paul: Well, we will see. 

Silvija: I have heard that there is some special hardware being produced for this. Makers like to talk about Rasberry Pie, and then there is this Arduino-chip. Teach us a couple of cool, nerdy concepts. 

 Paul: Ok. Let’s start with Arduino. They were some university lab, I think in Italy or Spain, which was loosing their funding, so they just published everything to make sure that it wasn’t lost. So it became an open source in different companies, so they can produce this small computer that costs a few euros. And you can make it part of a very simple design. It usually does not even have a radio, unless you plug in an extra board. A bit more expensive, is the 35 euro Rasberry Pie, and this is essentially the same as the computer that you use everyday. Just a bit less capable. And you have to plug in a memory card, a keyboard and a mouse. 

Silvija: And it’s small?

Paul: Yes, it’s quite small. And it allows you to put a computer in places where it wasn’t previously affordable. For example if you want to put one underneath your TV as a very common application for home use. I confess that I don’t have a smart home. I haven’t found the use case for myself, and my wife is a technophobe. 

Silvija: Must be an interesting marriage. 

Paul: Yes, let’s not go there. The Pie is also something that we use in a self driving car. We like to play with technology and motivate our consultants. We have a contest between our offices - Munich vs. Helsinki vs. Tampale (?), another city in Finland, to take a Raspberry Pie, put it on a remote controlled car and make it drive itself around the track. And this sort of stimulates us through competition to see how far we can push a very low end technology, since the whole car costs about 150 or 200 euro, in order to be ready when customers come forward to make something else that perhaps uses a camera for a different purpose. 

Silvija: Help us visualize something. So there is this tiny, little computer. Raspberry Pie may be 5 by 5 cm and 1 cm high. How do you connect it to the car? How does it read the world around it, and the device that it’s attached to? 

Paul: Well, it’s just a 5 by 5 computer board. It has a few USB plugs, like on your laptop, so you can plug in your keyboard and mouse that looks like the one you use to charge your phone. And you usually put it inside some sort of plastic box. At that price it doesn’t even come with a box, so you choose what you would like. I typically print the one that I am going to use. Then you have a lot of pins sticking out of the top, and you can attach things to those pins, for example motors, sensors and cameras. 

Silvija: These are the things that measure distances, or whatever you need to measure in order for it to do it’s calculations? 

Paul: Yes, so you go on a website, and say for example that for this particular experiment I would like it to have a camera, and I need a wider camera, so that it can see the entire road in front of my electric car. I also need a motor for the car, and steering. In our case, we are lazy, so we just buy an electric car and strip off the radios that come with it. 

Silvija: You just buy an electric car? 

Paul: Well, we bought a particularly fast one, so it’s fun. 

Silvija: I’m beginning to understand your wife. So tell us about your most interesting projects. Practical, concrete projects in IOT. Not necessarily so practical, but the most memorable projects. 

Paul: One that sort of sticks out, and got a lot of press coverage, partly humourous, because I instrumented our entire office with radios that I stuck on the walls, bluetooth beacons that are constantly pinging out of signal and depending on how strong the signal is, I will know if I am close to the beacon. As you walk around the office you could look at your phone, and see where you are, as well as see where everyone else is. In an open office, you don’t necessarily know that the person you want to talk to is over by the coffee machine, but then you can go and catch up with them at that moment. Or you can see a notification that something is happening, like if someone made a goal on the pool table, and here is a short video of it, or that there is food in the kitchen to celebrate something. 

Silvija: But it has a big brother connotation. 

Paul: Usually, but if you look at the website, it’s all open source. It’s not a commercial product. It was an experiment - VOR.space. You can see how it’s done. I’m anti Big Brother, and I don’t have any Amazon devices etc. in the house. And that’s one of the reasons why I am happy to be doing that for an office situation. I think that the information should stay locally. Because it allows you to be more comfortable with that. In the case of VOR, which is named after some Norse goddess of holmes or something, and I don’t even know how to pronounce it - we kept all the information inside the office, and we kept no records. You can summarize it as what happens in Vegas, stays in Vegas. There is no record, so it’s just what's happening here and now. You get a sense of what is happening in the office around you. And by not even sending it to the cloud, just having it transmitted to other people who choose to share their location, it’s really pro-people. 

Silvija: But at the same time, what you are doing, is doing a very interesting social and ethical analysis of how this technology should be used, for example in hospitals, where we really need to know where the nurses are, where the babies are, where the patients are and where the equipment is. And how to deal with that, which is perhaps a little bit more intimate or abusable information. 

Paul: Each case is different. In hospitals, there may be a need for records, at least having to do with the people who work there. That might be a useful thing, but you have to be very respectful and aware of the potential of abuse. We don’t want to create some dystopian Big Brother who knows if you are reading the wrong books. We are interested in using this in positive ways by sometimes intentionally limiting it, or limiting the scope of how far that information goes. You might process the information, so you don’t actually know. I can find out, for example, how many people are in this building just by listening to your phones, but I don’t know who you are, and that’s good. If I can find out that this particular room is heavily used, then that’s useful information. 

Silvija: And you don’t need a heavy text-stack to do that kind of stuff anymore. Basically, if you know how to read those phones, then you can get that information just with a couple of relatively small computers.

Paul: Yes, and I think that from a business standpoint you need to very much protect your ability to control your own service. If you want to build a service like that, you should not be too locked into a proprietary service, and with open source, or by doing a little bit yourself, and putting together a few pieces, that’s entirely feasible. 

Silvija: Which is why everybody needs to learn about IOT, as you keep saying. We are approaching the end, and I would love you to say a little bit more about IOT-service kit. And I will just state that we could have had a great debate, or conversation about the controversies. You touched upon who gets to control this, but there is also who owns the customer, where should the data live, and loads of other things. 

Paul: That is the meat of the matter, and inherent in how you approach the problem. 

Silvija: Exactly. Tell us about IOT-service kit. 

Paul: So it’s IOTservicekit.org, and you can download the entire thing. It’s not a commercial product. You can print out a map, and let’s say that you want to build a service for your office, so the map of your office gives you a game-like way to take 3D-printed pieces, legos or cards for less common things, and experiment in a gameboard type of format about the scenarios that your users go through to use that. It isn’t a highly structured thing, but a situation where you bring together a designer, a business person, an engineer or whoever is directly involved, but not more than five or six people, to try to simulate that service and go through these use-cases where Internet of Things could be useful. Typically when I come within ten meters of that bluetooth beacon, we know that, so that gives us the opportunity to change that information on a screen nearby, and these types of things. It won an award, and it’s good fun. We do that with customers, and for our own purposes as well. 

Silvija: I have a million other things to ask you, and I see that my boss here is looking at me, so we need to land. I asked you for a favorite quote, and I’m simply just doing to read it, because I love it: “There is no cloud, only your data on someone else’s computer.” Which is worth remembering. And then there is another one which is a bit bigger: “There is no faith, but what we make.” But that is the maker in you, right? 

Paul: We have to choose to make things that help people. 

Silvija: What is the most important thing that people should remember from our conversation? 

Paul: That it’s not difficult. These things are possible today. You should test the idea as quickly and as cheaply as possible, because you want to iterate through several variations to make sure that it has a human need behind it which is well served. And that there is a novelty factor that you need to get past to test it quickly. 

Silvija: And it’s worth doing things on your own and not just buying a completely done stack. 

Paul: Yes, and it might not be as expensive as you think to take freely available things versus using a commercial service. 

Silvija: Paul Houghton, thank you so much for coming all the way from Helsinki and teaching us about Internet of Things. 

Paul: Thank you. 

Silvija: And thank you for listening.