Arm Mbed

A short (and personal) history of the micro:bit

Just over a year ago, amid significant fanfare, the micro:bit was distributed to a million schoolchildren – that’s every Year Seven child in Britain – in a bid to tackle the technology skills gap and create the next generation of digital innovators.

Now, it’s well known that the micro:bit is based on an Arm embedded system. But what’s less well known is that the foundation for the micro:bit was laid nearly 20 years ago, in the midst of a drive to engage students with computer technology. (Can you see a pattern here?!)

At the time, I was doing outreach work with schools and colleges, but found that the institutions were struggling with tools that were too professional – too detailed for the students that were trying to use them; these were complicated things that took half a day to download and needed a ton of technical knowledge just to get started.

In December 2005, I submitted an application to the Arm patent committee to cover some of the ideas I had around compilation and hardware programming. The application was rejected, but as a result, I got talking with Simon Ford, who was doing a similar thing in universities – and encountering the same issues. We discussed providing tools that were hosted online, rather than requiring download, and floated the idea of a software library that took detailed technical interfaces and made them simple enough for non-experts to use. Mbed was a product of this.

Some “founder photos” that were used may times in many publications and blogs as Mbed gained traction. Looking a lot younger back then!

By the time we released Mbed in late 2009, the platform had been in a closed beta for over a year. We’d styled it as a Rapid Prototyping Platform, providing a small hardware board, online compilation tools, and a software development kit that made using the technology intuitive, and had managed to build a small but thriving, self-helping community around it.

These initial beta-users were like-minded developers; early adopters who understood that we were trying to do something different. Their knowledge, throughout this prolonged period of user testing, helped refine the developer experience that, ultimately, set Mbed apart.

We ran numerous Mbed workshops at schools and universities, and watched as people tried to use the platform, often in ways we’d not envisaged.

Working with a year 11 student at Linton Village College during the early user experience workshops. Students had great ideas of things they wanted to make, and Mbed was a way for them to not have to sweat the details.

Arduino – a similar project, aiming to create simple, low-cost tools for non-engineers – was happening at the same time, and trying to solve similar problems, but was directed at artists who wanted to apply technology in art. We had no specific target audience, other than bright sparks who may never have used electronics or software, but could easily envisage the advantages that the technology could bring. We wanted to create an off-the-shelf system – compatible with industry standard software, tooling and methodologies – that would give them the tools they needed, without a huge investment in time or money.

It was a popular idea, but it was difficult to outline the business case for it. Eventually, however, Simon managed to persuade Mike Muller that it was a cause worth making a small investment in, just to see where it might lead. Fifty per cent of both mine and Simon’s time was allocated to the project, and we started building the first proof of concept.

We traded a cupboard for a bottle of wine and squeezed in desks for me and Simon, and a hot desk for the small band of contractors and collaborators we were assembling. By the end of the year, we had a fully-functioning platform: hardware, software, the lot.

At the ESC West Coast in 2008, we had a cork board on NXPs stand and Simon attended to socialise the Mbed concept around the industry.

At TechCon 2008, Simon showed it to NXP, who were keen to get onboard; we spent the next year doing user focus groups and honing the product. By the time TechCon 2009 came around, we were ready to launch. As with most MCU development boards, NXP expected to ship between 500 and 1000 during the life cycle of Mbed LPC1768; to date, they have shipped over 140,000 – far in excess of their next best-seller.

After the successful launch of Mbed at TechCon 2009, an Mbed shaped cake was commissioned and cut for all to enjoy in ARM1 atrium.

In 2012, we brought in a second silicon partner – Freescale – and did an open-source release of the Mbed SDK, which had spent three years as a closed-source product. Over the next year, the phone just didn’t stop ringing with calls from silicon partners who saw the potential of this open source, multivendor platform, and wanted to be a part of it.


But what has this got to do with the micro:bit?

Well, in summer 2014, the IoT business unit had just formed, and Mbed was on a mission to connect the world’s devices. Just 36 hours before the proposal deadline, we heard that the BBC was looking to commission a product for day-to-day usage in schools. At that point, they were expecting to make another microcontroller development board, but a bunch of us sat down around a whiteboard … and quickly whipped up a proposal to convince the BBC that the next wave of technology they should be promoting to the nation’s students was all about device connectivity, and they could easily build on the work we’d already done with Mbed.

We drew heavily on Mbed’s usability. As far as the computer’s concerned, Mbed is just a flash drive. That makes it brilliant for schools: students are rarely allowed to download software onto school machines, but they can use a flash drive. In addition, drag and drop programming makes it easy to pick up and use, and it works on Linux, Mac or Windows, too.

Proudly holding one of the first prototyped boards.

The fact that it was a working product made it an easier sell as the basis of the micro:bit: a lot of the potential teething troubles had already been ironed out, making it easier to integrate with Microsoft’s user-friendly graphical development environment.

And the rest, as they say, is history. The micro:bit is now in use in over 70 countries around the world, and the goal of getting them in the hands of 100 million young inventors within 10 years is looking very likely to happen.

In an uncomfortable comparison with monkeys, typewriters and the works of Shakespeare … who could have predicted that a few men in a cupboard could come up with something as useful as that?


Want to know more? Get creative with the micro:bit at microbit.org.