The humble Raspberry Pi. A pint-sized fruit-inspired computer that was designed to increase childhood interest in computer science… that’s now used by millions of children, academics, hobbyists and businesses all over the world.
It’s been a decade since Eben Upton (pictured right) – a British technical director and ASIC architect for Broadcom – founded the Raspberry Pi Foundation, the charity behind the Raspberry Pi computer, and 7 years since the first sale of the Raspberry Pi was so popular that it crashed the website that was selling it.
We were keen to share the story of their success and how the organisation has evolved over 10 years to become one of the most important developments in enabling both industrial businesses and equal education, so we caught up with Eben to find out more.
Alice: What inspired you to develop the original Raspberry Pi computer and foundation?
Eben Upton: About a decade or so ago, I was involved in interviewing prospective Computer Science undergraduates at the University of Cambridge.
The numbers applying compared to a decade before in the mid-90s had really declined. And we had this depressing realisation that there just weren’t that many people interested in it anymore.
We’ve sold 25 million Raspberry Pi’s now so it’s a slightly different scale to what we imagined
Historically we haven’t really taught computer science in schools to any great degree, so we were getting our applicants from home computing in the hobbyist world.
As home computing as a hobby declined during the 1990s, our supply of prospective undergraduates dried up. And really, the idea of Raspberry Pi was to see if we could reboot the home computing hobbyist world that used to exist in this country – and if we could reboot that, would it lead to an increase in the number of applicants.
Alice: So quite a surprising and positive outcome considering how big it’s gotten and how it’s affecting education now!
Eben: It’s enormously bigger than we were expecting it to become. We thought if we got a thousand machines into the hands of the right thousand people, then we could make a difference.
We’ve sold 25 million Raspberry Pi’s now so it’s a slightly different scale to what we imagined.
Alice: What are some of the challenges that you experienced when scaling up Raspberry Pi?
Eben: What’s quite unusual is that we’re a charity, so we have a different set of constraints than what confronts most commercial organisations.
One of the biggest ones of course is that we can’t raise money, so we’ve had to build this business without going and raising external capital. We had a little money upfront – about $200,000 – and that was enough to build 10,000 units of the product.
For our initial idea of what the scale might be, that was fine. But as we started to talk about the product in public, and appreciate that there was going to be quite a lot of public interest in it, it became apparent to us that we were going to overrun our initial scale and therefore the supply of capital wasn’t going to be enough and we had no way to raise additional money.
So the biggest challenge that confronted us early on was how to grow without increasing the capital intensity of the business and the answer was to become a licensing company.
We design the technology and then license it to partners who commit the capital and build the product. That means that we receive a royalty for the technology which allows us to focus on what we’re good at.
Alice: And how have these challenges evolved over the last decade?
Eben: Early on at least, it was very challenging to build anything at the target price, because we were building such low volumes.
We still think of ourselves as a toy company to some extent
We got quite a lot of support from vendors, in terms of not being penalised for being so small, because we are a charity – and then later on the business became large enough that it could stand on its own in terms of its relationships with suppliers.
Cost is always a challenge because, as we become more ambitious, we want to pack more into the platform and then it becomes a constant game of trying to recover and scrape together cost savings in bits of the platform in order to get results in other areas.
Over time it’s become more about continually chasing cost savings in order to add some functionality, whereas early days it was about whether you could even build this thing at all!
Alice: Although it was originally designed as an education tool, Raspberry Pi’s have started to be used more in industry environments. What are your thoughts on that?
Eben: We still think of ourselves as a toy company to some extent – but what’s interesting is that a lot of the things that make Raspberry Pis a good educational tool also make it suitable for industry.
It doesn’t consume a lot of power, it’s low cost, it’s robust, it’s relatively high-performance and has a lot of interfacing capabilities.
[Raspberry Pis] are mobilising energy and creativity on a much broader scale that we could ever believe possible
Looking back it’s unsurprising but if you’d told me 10 years ago it would happen I would have been amazed.
Of course, it also challenges us to make our products more robust for industry, which also make it more survivable in an education environment. So they feed across into each other which is nice.
Industry sales count for over half our volumes, with over 3 million units last year going into industrial applications. And sales in industry, just like sales to education, generate money which we’re then able to employ in the service of the charitable mission.
Alice: When it comes to your favourite Raspberry Pi projects, or success stories, have you got any that stand out?
Eben: One of the exciting things about Raspberry Pi is the businesses that have sprung up around it.
For example, there’s one in Sheffield called Pimoroni which started off with a laser cutter and now employ 40 or 50 people. We also have some guys up the road in Cambridgeshire running an organisation called The Pi Hut which has grown enormously on the back of selling Raspberry Pis and accessories.
There are also some fantastic hobbyist applications of Raspberry Pi. You have people sending them up on weather balloons and taking pictures from the edge of space.
There’s another I love that’s by a Japanese guy who made a cucumber sorter for his parents who are cucumber farmers. He used Google Tensorflow image processing to look at the cucumbers and classify them into one of many different classes – helping his ageing parents to run their farm.
When Raspberry Pis get into engineer’s hands, they invent in ways that’s way beyond what we could’ve ever achieved ourselves on all sorts of levels. On a charitable level, like the clubs that are run around Raspberry Pi, they are mobilising energy and creativity on a much broader scale that we could ever believe possible.
Alice: What do you think is the biggest factor behind the success of Raspberry Pi?
Eben: I think it’s just that there was a latent demand. Often when you start a new company that sells a radically new product, you need to educate customers about what your product is.
The world is full of these reservoirs of demand that are not being fulfilled
I never got the sense that we needed to convince people that they wanted to buy it because people already knew that they wanted a product like Raspberry Pi but couldn’t quite put it into words.
The world is full of these reservoirs of demand that are not being fulfilled – and one day you might be lucky enough to stumble across one and I think that’s what happened to us.
Alice: When it comes to the aims of the Raspberry Pi Foundation, what are you biggest hopes for this in the coming decade for what it will do for children across the world?
Eben: I think we are seeing good progress in the UK in terms of our initial very narrow goals. The number of applicants to Computer Science at Cambridge has gone up by around a factor of six, so we went from having 600 applicants in the 90s to 200 applicants 10 years ago, to 1100 applicants last year.
But I think there’s a lot of unfinished work in the UK. In particular we’ve just become involved in a government programme to train teachers.
Having good computer teachers provides uniformity of access to computer science across geographic location, socio-economic background and gender that you’re not going to get if you rely entirely on self-directed learning.
Alice: So having that uniformity in education, what kind of outcomes would you like to see?
Eben: An engineering community that looks more like society.
Go into any engineering company and you’ll see it’s not representative in any respect of the rest of society, or the rest of the country.
That’s not really very good, and it sucks for the country as there’s talent out there in other people that isn’t being utilised. And it sucks for those people too because, through the operation of the system, they’re being denied an opportunity to get involved in something that’s extremely rewarding in all respects.
and featured on the Raspberry Pi blog
Lots of the effort we put in now is about making stuff relevant. Computing is not just ‘further further maths’, this is not just an abstract set of operations that you perform, it’s a tool that you can use to accomplish other goals you have.
For example, to build an art installation, measure rainfall, take photos of animals in your garden, make a light-up dress, measure how fast your bike is going – none of those things sound like computer science but they’re all things that computers can help you do – and that’s the message we’re really trying to get across.
Header image credit: Tech Crunch
If you’d like to discuss your startup or project, get in touch with Simpleweb today.