Amazing new little device, and with it's much more powerful MCU (nRF52-series Cortex-M4 rather than the nRF51 Cortex-M0+) with 8x the RAM you can now even run small Machine Learning algorithms on it. Here it's responding to my voice: https://www.youtube.com/watch?v=fNSKWdIxh8o&feature=emb_titl...
Thanks! We got approached by some of the people in micro:bit to see if we could get this to work on short notice, and they sent some samples out last week!
Is this achieving the goals of getting computing power into hands of kids ? I know they gave away a million a year or two ago, but my kids were more interested in raspberry pi as a platform - the micro-bit just kind of did not have the ... marketing ? appeal ? obvious way in ?
I like the idea - I can see some kind of path to free/open IoT devices and some standardisation. But ... is tehre some call home feature that says 'I have been turned on 5 times'?
Is this effective ?
I know I sound like an old moaner - it just is an important area - how do we measure if we are succeeding?
My kids love microbit - and have been hacking stuff on the platform since they were 7 or 8 years old.
They’ve played with raspberry pi too, but Microbit is just so immediate. You write some code (or drag some blocks about), upload and it just works. It’s simple, understandable and quick. It’s also bridging the gap to learning some basic electronics too - building small circuits on a bread board with a mictrobit controller.
There’s something about microbit that just sparks my kids imagination - they’ve used it for school projects, homework assignments, they make little robots, figure out how they can automate tasks like watering their plants. All things they could do on Raspberry Pi of course, but still…
I see my kids playing about with mictrobits and I get a really sense of the same joy I had messing about with computers and electronics in the 1980s.
Many of the resources we make at Raspberry Pi, actually target the micro:bit[1].
We recently worked with UK Scouts, and again used the micro:bit for many of our resources [2]. They're just easier devices to use, especially for some Scout leaders, who are not technically experienced.
I think of these devices as part of a long term, broad based parenting strategy aimed at opening mental access to a variety of human activity that includes but is not dominated by technology. The measures of this are very person-specific. It could be engagement, attainment, development, etc. I don't think there is a measure for all of "we."
As far as I'm aware there is no "call home feature" and I am very glad that there isn't as my goal is to clear many paths for the kids to choose from and not excessively encourage any particular one. If one were to have a measure, you can connect MakeCode to a GitHub account and track commits. If that became the measure, I'm sure the funnest kids would find a way to break it with negative 2000 lines of code [0] or the like.
Having tried Raspberry Pi, micro:bit and ESP8266 with my kids at various points I've found the micro:bit and MakeCode package to be at a sweet spot of simplicity and "batteries included." It contains many features like Bluetooth transfer of code that is brilliantly easy and enough sensors and lights to see something real without much yak shaving. There is a sufficient ecosystem of micro-bit slotted servo drivers and complete kits to fit different learning experiences.
If a person is interested in more sophisticated programming then a full Linux machine is great. If they want to sprinkle sensors everywhere then the Arduino universe is great. For someone who is curious but not committed then the micro:bit is an accessible path to develop a budding interest.
I've got quite a bit of experience of doing coding stuff with kids - I set up the CoderDojo at The National Museum of Computing in the UK (paused at the moment), have run a Code Club at our local library, have two kids myself and am a Raspberry Pi Certified Educator. I've done quite a lot of Microbit sessions for kids and a handful of Raspberry Pi ones.
I really like Microbits for all sorts of reasons and they fill a very different place in the ecosystem to a Raspberry Pi. They are cheap and have a low entry threshold - I can get a kid doing interesting stuff on a Microbit in a matter of minutes, but there's still interesting potential if you want to go in an electronics direction with them. They are also fairly indestructible so you can let kids tinker with them more. There's not much you can do with a Raspberry Pi that you can't do on other computers without buying something like a SenseHAT or going down the electronics route. It's hard to do electronics with kids without close to one-to-one supervision.
I can see for older kids with tech-savvy parents who are willing to invest in electronics components etc. there are more possibilities with a Raspberry Pi and there is obviously lots of fun stuff you can do that you could never do with Microbits, but I definitely don't see them as direct competitors.
The other thing I will say is that I have noticed a bit of a personality thing too - some kids get super-excited by making the LEDs on a Microbit light up and so on and some really don't care. Different types of coding definitely appeal different to different children.
I should add that it's actually the speaker and microphone that I am most excited about in terms of the potential for interesting new projects for kids. Crocodile clips and headphones were quite cumbersome, especially for kids. Although I might have to get used to more noise in our sessions!
In terms of rapidly getting up & running with projects, I think micro::bit fills a nice niche in between an arduino and pi. You can have a light blinking in a few minutes on an arduino, but not much more without add components. A pi can obviously do tons of stuff without any additional hardware, but you have to boot a full OS to get there. A micro::bit gets you up & running about as fast as an arduino but you can code more meaningful projects without any additional hardware.
It's a very versatile device. I paired mine with a LoraWAN Hat and it sits outside in my garden, in a waterproof (unventilated) case measuring temperature and light. It's worked reliably through heatwaves and cold, wet weather and only stops working when the batteries die. It's very tough.
Cool! Refreshing with new hardware capabilities, and it seems they thought it through with the addition of a universal (yeah ...) hex format to support both v1 and v2 platforms from a single file.
The new nRF52833 MCU [1] seems pretty capable and certainly inline with what is being used in industry for suitable applications (it's an Cortex-M4 64 MHz, 128 KB RAM, 512 KB flash, Bluetooth) but it doesn't seem to have hardware support for any ML-specific stuff that I could see.
Oh, and there has been a keming accident on one of the images that is a bit unfortunate, the word "MakeCode" looks a great deal like "MaleCode" which is perhaps not optimal in their context [2].
Glad you like the Universal Hex format! And thanks for the headsup on the SVG - we noticed it earlier and there's a fix on the way through the site build(though whether you get 'Male code' or just bad kerning seems weirdly system dependent). Fix via https://github.com/microbit-foundation/dev-docs/pull/168 :D
Thankfully not - the file gets larger but as there's an MCU managing the writing to flash, it selects the right binary for the appropriate device.
There are some limitations around ensuring backwards compatibility: if you want to write a program that runs on both devices, you have to adhere to the constraints of the smaller one - but I think it's a fair tradeoff!
It has a 32-bit FPU, with single cycle add and single cycle multiply; it also supports 16 bit float storage format. It lets you run small models, but note that you can't fit big ones in flash anyways.
There are also some DSP instructions, but it's only 2x16 bit mulitply and 4x8 bit saturating addition, so not so much use.
Espruino (a JS interpreter for MCUs) runs really nicely on this too now - you get wireless programming and debug, and a full Bluetooth LE stack (you can run Web Bluetooth on the device and use to control other consumer electronics).
I was just about to comment saying that it was nice that they mentioned Espruino in the article and it would be great to see an Espruino build without the limitations of the old device. Great to hear you're already working on it!
Any idea which Zigbee isn't mentioned in the press release, when it seems to be available in the new SoC - would it be problematic to make use of it in the micro:bit?
I think there were just a bunch of other features they wanted to give exposure (ML, speaker, etc), and the kid-friendly tools out there for it won't support Zigbee.
However I bet Nordic's examples for Zigbee on nRF52833 would work pretty much as-is (maybe just changing a few pin names) so you could get up and running pretty quickly.
I'm very excited to see nrf52833's sweet new Bluetooth 5.2 capabilities made accessible to makers & learners. Would that more devices of all manners try to offer this kind of up-to-date bluetooth support!
They are great fun. Played around with the previous generation for use as a bluetooth beacon amongst other things. I'm really excited by the idea that this new generation could be used to recognise audio!
Great announcement. Also great that the LED array in version 1 was empirically a success.
As frequently as Raspberry Pi devices are mentioned in comparison, there also mbed-style [1] inexpensive boards which support the "online compile and download to a virtual USB drive" model, and offer free RTOS support. (Perhaps these too -- or variants -- could hit a sweet spot in educational contexts which Raspberry Pi could miss.)
Edit: micro:bit is included there as an mbed-compatible device: https://os.mbed.com/platforms/Microbit/ . So presumably one can also practice RTOS for IoT on a micro:bit .
Very nice. Does this mean that micropython will be able to do bluetooth now? The extra flash memory will be a boost for sensor-based datalogging, if only you could write to files in append mode. Is the updated micropython source tree for this available anywhere yet?
Headsup that the location of CODAL (the underlying layer that supports the micro:bit) is going to move today as part of going private to public so there will be times when that repo doesn't build (or at least, docs are out of sync with where the repos are) at some point during the day.
The MicroPython build has SoftDevice in it, including a Bluetooth bootloader so you can flash MicroPython programs using the mobile apps (betas being released today, too, see https://tech.microbit.org/software/beta/). However, we want to do the BLE API design in the open for how MicroPython will be a GAP Central or GAP Peripheral. Of course, as the code goes upstream it should also be possible to use other BLE stacks and existing MicroPython APIs, but for now our focus is on the classroom environment and getting consistency in experience on MicroPython and BLE flashing/update.
In the classroom we strongly expect people to continue using micro:bit Radio for user programs because of how easy it is to use and the way it allows micro:bits to communicate with each other.
The micro:bit radio is fantastic for most classroom cases, but I'm doing a bit of teaching around data, and micro:bits really struggle with that - between the lack of append mode for logging the data, and getting the data off them for analysis on a real computer, it's been pretty frustrating - hopefully this new version will help :)
Very cool. Does it include battery charging/management yet? That's a complicated thing to make and it seems like a pretty important feature if you want to allow people to make cool battery powered devices.
No battery charger on board, but we’ve done a lot of work on power management so you can sleep with very low power consumption (>100uA) on USB - which means you can use USB battery packs. AAAs work really nicely in classrooms and for teachers (easy to buy, clear what you need, low cost batteries are safe, also familiar for the existing 5M users), and have a lot going for them in terms of safety for open electric devices used by novices! The AAA batteries can also last a lot longer now if you design your code to be lower power.
Most USB battery packs will go to sleep themselves though when you use that little current, and turn off their output. But some have a "low power" option that can be activated to keep them on.
You can put it to sleep now from software, and also you can safely connect a LiPo battery (not only 2xAAA). There's no LiPo charging circuit on it though.
WiFI would have been awesome! I love the accessibility of this thing, and it’s a great place to get started for hardware enthusiasts since you can drag drop in the browser.
You could totally build a guitar tuner with the latest version
I agree that wifi would be great, but given the target market is schools, and that schools run enterprise wifi, it would be a lot of trouble (for both the teacher, and the IT team) to get it working, and probably lead to disillusionment with the device, so I respect the decision to keep wifi off.
There is an ESP32-based microbit 'clone' called the BPi:bit (BPi = BananaPi) you can get off AliExpress that includes wifi (plus RGB LEDs), which might be worth a look
Use microbit fir my weather station but as said you need other parts. It is easier than pi/arduino or the fpga I used as well. Not sure I will invest (my time) more though.
Isn't this a competitor with the Raspberry PI - who had the same goals? And this was sponsored by the BBC (i.e. via licence payers money). And free devices given to children.
Shoudln't the BBC worked with the Raspberry PI foundation (charity) and given Raspberry Pis to children? Also the investment spent in making the PI not 'overwhelming' - which my 9 year old says it isn't (btw).
No, it's not. This is specifically targeted at kids to learn simple programming. It does not play in the same league as the Raspberry PI.
Programming this thing is amazingly simple. You use a browser based visual editor (at https://makecode.microbit.org/ ), and when finished you download the file and store it on the micro:bit just like storing a file on an USB drive. It will start automatically. Really simple so children see results as fast as possible.
You should compare it to programming an EPROM with your 8-Bit computer, because the micros-bit obviously does not have a screen, keyboard or connection for external storage.
I know what you mean, I still have my Commodore 64 and it still works, even the Competition pro does. And I still have S.E.U.C.K., which can be seen as the ancestor of Scratch, on floppy disk.
But that's software not hardware. That could run on a pi. A PI has a 'usb drive' and you could code the OS to 'start the file' as soon as it is downloaded.
Were PIs not conceived to be targeted at kids to learn programming ?
You can't run Pi on batteries (at least not for long). If you want any input or output device, you'll have to buy it, which makes it difficult/expensive to embed in projects (eg https://makecode.microbit.org/projects/toys).
This is cool and all but let's take a step back.. what can this thing do that much more powerful computers and smartphones can't? If the answer is "nothing", perhaps they are just producing unnecessary litter?
It’s a platform for kids seven and up to learn about computing and electronics. You can buy a classroom set for about £140, less than almost any single smartphone and far better suited to building things in both software and hardware.
The point of the micro:bit is education, which requires it to be simple. The Raspberry Pi was made for education, but is completely overwhelming for young kids.
The micro:bit is extremely simple, just a few buttons, pins that can be used easily with crocodile clips, and a small grid of LEDs that makes it easy to draw primitive graphics.
We gave our daughter a micro:bit when she was nearly five (she is six now) and she loves it. The drag and drop programming that MakeCode offers is something that someone of that age can use to make very simple programs (push this button -> show an emoji or write your name). Seeing their program on real hardware then gives them a lot of satisfaction.
tl;dr: this is not competing with your average Linux-running ARM board or even other microcontrollers. Their goals is offering something that is fit and safe for education.
