Input sound file:
10-16 pm Session 1_1002.MP3
Session name: Raspberry Pi – a Lego brick for the internet of things
Christ Albrecht 00:00
Back to the stage, my colleague Stacey Higginbotham, she’s going to be talking with Evan Upton, the co-founder of the Raspberry Pi Foundation. They’re going to be talking Raspberry Pi, a Lego brick for the internet of things. Please welcome our Raspberry Pi Panel.
Stacey Higginbotham 00:20
Alright, well thank you guys. This is really exciting, especially for those of you who really like to play. I like to play with all kinds of things, hardware, putting stuff together, making things happen. It’s fun. And I was really excited about having Eben here because you, through Raspberry Pi, are bringing this sense of fun to kids everywhere. So I thought we’d start with– hey, thanks for coming. And also getting you to tell us a little bit about why– we should say what Raspberry Pi is first for those of you who don’t know, and then why you are doing this?
Eben Upton 00:59
So Raspberry Pi is a 25 dollar credit card sized Linux PC that we developed in Cambridge to try to get kids programming again. And this kind of grew out of a really bad experience for the group of us the university had about a decade ago, where we looked around one day and suddenly we found– so the University of Cambridge is one of the premiere engineering schools in the world. It’s definitely the premiere engineering school in the UK, and we looked around and we found that we were struggling to find enough young people who wanted to come to the university to study computer science and engineering. And this was against the background of the mid-90s, having been overwhelmed with applicants, and finding that when those people arrived, pretty much the first thing you had to do was convince them that there was something they didn’t know. These people came in the door with so much computing knowledge, so much computing experience that they’d acquired on their own at home. We were spend the first few weeks beating them over the head before they were even prepared to admit that there was anything about computers that they didn’t know. And within a decade we’ve gone from that wonderful environment to an environment where a) we were struggling to get people in the door and b) when people did come into the door, they were very bright kids, they’d have like the level of experience they had in the mid 1990s.
Stacey Higginbotham 02:12
Why was that?
Eben Upton 02:14
We have a hypothesis, and Raspberry Pi is still– we only still have a hypothesis. And our hypothesis is that people of my generation got our experience with computers from these machines that we would have in our bedrooms. That we would have a Commodore 64 or TRS-80, I have to localize the computers I talk about to the country but, some people would have a Commodore 64. The good thing about Commodore 64 is that they sold everywhere so people would 64s or TRS-80s. In the UK people might a BBC micro-computer, that was what I had. And you might buy those to play games on or you might buy them to do your schoolwork on or your parents might buy them for you to do your schoolwork on, but they all had that one thing in common which is you turn them on and they go beep. And the first thing you can do is program them. In fact if you want to do something other than program them, the first thing you have to do is to choose not to program them, so the default choice of these machines is to program. And what I mean is almost everyone had that experience of being able to write at least 10, Print [inaudible], 20 go to 10, fill the screen up or something filthier than that. And then go into your local computer store and type it into all of the machines and then run out of the door. That means everyone my age in the developed world who had any interest in computers at all had a chance to discover it. And then had a machine just sitting in the corner of that room with a blinking cursor, just tempting them to get more and more involved. I think a lot of us owe our careers to that very early experience of having a very programmable piece of hardware.
Eben Upton 03:37
And then that hardware went away. That hardware was replaced in the 1990s with much more user friendly, much more usable hardware. Which is great for the average user, but what it does mean is that this side effect of the unusability of 1980s computers, that we generated very large numbers of engineers. That side effect went away. And so 10 years after those machines went away, our children stopped being tricked into learning to program. Our supply of 18-year-old undergraduates, plenty of undergraduates [inaudible].
Stacey Higginbotham 04:06
And so you guys in February of last year, announced the Raspberry Pi. And you expected the goal was to jump-start this programming generation again. And you expected to sell ten thousand of them you said?
Eben Upton 04:23
The scale of the ambition was really shockingly limited. So we were trying to solve a very problem from one university of one country. The decline in the number of people was precipitous in relative terms but actually very small in absolute terms. We’d gone from maybe 500 applicants down to 250 applicants. So really the thinking was if we could get a thousand machines into the hands of the right thousand children we could reverse this decline. So really Raspberry Pi started off having these very small volume ambitions.
Stacey Higginbotham 04:54
And instead you’re up to two million now.
Eben Upton 04:57
We are closing in on two million.
Stacey Higginbotham 04:57
There was a shortfall when you guys first launched and to me this is exciting. I’m a parent, I have a seven year old daughter and I want her to play around with hardware, learn to code, I don’t code, I feel kind of sad about this. So I–
Eben Upton 05:17
Should get a Raspberry Pi.
Stacey Higginbotham 05:18
I should, I have an Arduino I’m sorry. I know! Shocker. But when I look at this I’m really excited by the opportunity but I’m also wondering as a parent, as an educator, how do you get kids involved? You don’t just slap a Pi on their desk and you’re like, “Go! Make it do something.”
Eben Upton 05:40
I think what we’ve learned over the five or six years we’ve been doing Pi and the nearly two years we’ve been shipping Pi is exactly that. That just putting a piece of hardware in front of maybe one percent of children, simply putting a piece of hardware in front of them is enough. One percent of children will find that sufficiently challenging and interesting and intriguing that they will go and learn how to install an operating system on it. If you’re going to get beyond the usual suspects, if you’re going to grab to maybe 10 percent, I think 10 percent is maybe a reasonable target. 10 percent of children having a really decent amount of knowledge of how to code machines. You have to make it a little bit easier for them. You have to smooth the first little bit of that learning curve. I think if you can make the first hour, the first day straightforward, enjoyable, then I think for a lot of people that’s enough. But there are a lot of people who will stumble at that first hurdle. Because you have to remember a PC, many children have access to PCs. Children do tend to have access to PCs in the school, many of them have access to PCs in the home and it really is only 10 minutes to and download Python and install it on your PC. And there is already good evidence that children fall at a very very small hurdle and it’s getting rid of that as much as anything else that we’re trying to accomplish on Raspberry Pi.
Stacey Higginbotham 06:59
The interesting thing that I think a lot of people talk a lot about probably much to your chagrin is all the other things. Like the adults using Pi to build supercomputers and crazy connected devices, did you expect that?
Eben Upton 07:15
We really didn’t expect. We were very focused on a very small, very targeted intervention to try to improve this educational problem we have. So there were two things we didn’t anticipate. We didn’t anticipate the adult hobbyists who were going to use the Pi to have fun, and we didn’t anticipate the commercial uses. So we didn’t anticipate that people were going to start to use the Pi as an industrial computer, they were going to, try and start to use it as an internet of things, as a node in an internet of things. All of those things were a surprise for us. And to some extent while we’ve sold many more units to children, we think we’ve sold three or four hundred thousand units to children one way or another. So we’ve sold many more units to children than we ever expected. The bulk of that upside, the bulk of that surprise for us in terms of volume has been driven by these markets that we didn’t realize existed.
Stacey Higginbotham 08:05
What might be next for you guys, especially given that there’s this whole market, how could you exploit that?
Eben Upton 08:13
I think we have a perception particularly in the IRT space, we have a perception that the Pi’s form factor is engineered for this educational market, it’s engineered to be a great little computer for children that you can stick under your television and use that to code on. It isn’t actually designed to be embedded into stuff. And we see a little bit of evidence of stress, of people using the Pi and finding it a very high performance, a very cost competitive platform, but looking at the form factors and say, “If only that was a little bit different.” We have people get in touch and say, “Could I buy ten thousand Pis from you without the connectors?” So there is a little bit of evidence that people don’t want the form factor. So one thing we were considering doing over the next year is maybe trying to address that form factor. Trying to provide an alternative form factor which might be a little bit more suitable for people that are trying to use it as a component rather than as a physical product.
Stacey Higginbotham 09:06
Would that be like an Apple Pi or a Blueberry Pi?
Eben Upton 09:08
Yeah, we have to be really careful about one of those two words.
Stacey Higginbotham 09:12
Oh, yes, yes you would. I was just thinking of the most delicious pies.
Eben Upton 09:20
Interesting that it’s extremely hard to make a tasty raspberry pie. They melt away to nothing. So what you actually do is, the way to make a good raspberry pie is– well just in case because people do sometimes make us a raspberry pie if we go for a visit somewhere. People do make us a raspberry pie. The trick is to make a little custard pie with raspberries on top. So a raspberry tart with a little coolie.
Stacey Higginbotham 09:46
Eben Upton 09:49
That’s going on my writer.
Stacey Higginbotham 09:52
So there’s the Pi, there’s Arduino, there’s Beaglebone and there’s a bunch of hobbyist computers out there and Intel just did their Galileo board, so I’m curious how you see– is it competition? Is it just more fun?
Eben Upton 10:16
I think so far it’s more all of those boards add something new to the space. It isn’t so much that we’re fighting over a fixed size pie, but that we are – the puns write themselves – but we’re not finding a fixed size pie. All of these platforms bring new things. Galileo is a great example of that. Which brings XA6, the ability to run XA6, the XA6 instruction set in an Arduino like mobile form factor, is a really useful new thing. So all of these things move the market forward. One of the things I think very early on there was a feeling that we would maybe take market away from Arduino. I don’t think that there’s any real evidence that that’s happened. I think the sorts of things that people can do with Pi and the sorts of things that people can do with Arduino, there’s a little overlap so maybe we’ve taken away ten percent of Arduino’s potential applications but by popularizing the idea of cheap small board computers we’ve probably generated much more additional volume for Arduino. And certainly we see that with new things coming into the field. They just increase the amount of excitement around these IOT platforms.
Stacey Higginbotham 11:23
What are some of the most exciting or fun projects that you have seen?
Eben Upton 11:28
So I’m a child of the 80s. So I’m a real space cadet. The ones I like most are the space ones. So there’s a chap in the UK called Dave Ackerman who puts these things on the weather balloons and sends them up. We have a camera add-on. We don’t generally do add-on boards for the Pi but we did that 25 dollar 5 megapixel camera board. And Dave has been putting these on weather balloons which is kind of fun because he’s filling up this giant balloon with hydrogen. Because you want it to go up very high.
Stacey Higginbotham 11:58
Oh, so it might explode.
Eben Upton 11:59
Which is what’s cool about this. So if you’ve ever seen The Prisoner. If you want these things to go up nice and high you want to fill them so they’re marginally buoyant at ground level. And so the thing just wobbles up into the atmosphere. It’s a bit like the Prisoner. If you get it wrong– you get this big sphere full of hydrogen bouncing across the field, and there’s always that worry that over the next hedge there’s somebody having a barbecue, this big sphere of hydrogen will bounce onto the– but yeah, so he sends them up and then he takes these pictures and the wonderful thing about that is he’s 40 kilometers up. And you might as well be in space, in terms of the view you might as well be in space. And he’s doing this on a budget of two or three hundred dollars. Which means every primary school in the world, every primary school in the developed world can have a space program. And take these pictures and send down these pictures from space.
Stacey Higginbotham 12:44
I never thought about it as a space program. That’s far cooler than the catapults we made, and those were actually pretty cool.
Eben Upton 12:49
So a lot of what’s going on is we came from this very narrow world, trying to get more people in programming. But I think we’ve ended up in a space where, on the education side, a lot of what people are doing with Pi has brought a STEM mission. People are using the Pi in biology experiments, people are using the Pi for physics and climatology and all these things. There are some wonderful kids in the UK who are doing pollution monitoring. They’ve build a pollution monitoring shield, an add-on board for the Pi that measures air quality. Did a kickstarter and ended up deploying these all around the world. So it’s those kind of things that really take the Pi and use it to get kids– I don’t really mind if we get kids excited about programming or if we get them excited about climatology, as long as we get them excited about STEM subjects and that’s what we need.
Stacey Higginbotham 13:33
It’s true, and away from drugs. Alright, Eben thank you so much, I really enjoyed our chat.
Eben Upton 13:40