Rise of the Teenage Programmer ???

What a load of political guff. It ain't new, it's ancient history, but there's maybe a bit of political capital to be gained by reinventing the concept.

When the BBC micro was first introduced to schools in 19 oatcake we had Primary One kids (5 yrs old) "playing" with floor robots to make them negotiate mazes.They had to write simple programs and then input them via push buttons. Just simple commands like "Forward 10: Turn Right 90: Forward 30: Stop. But the basics of algorithmic programming.
We even had the older kids getting a pair of floor robots doing formation dancing for a parents evening. We also had a robot grab-arm that our top class could program to pick up and drop things.

 

Very true but after the "heydays" of the 80's, things went downhill.
Of course it does not help that in the UK and a lot of other countries, any I.T. at schools consists of learning how to do spreadsheets and if you are lucky, a few macros in VBA, hardly exciting stuff so many teenagers simply abandoned I.T. (and electronics).

However since the Raspberry Pi, the Arduino and a plethora of other mini computers became available, there has definitely been an upswing in youngsters becoming interested again.

As a side note, The Raspberry Pi was first manufactured in China, but now all manufacturing has been moved to Pencoed in Wales and is one of the biggest employers in the area.

 

Are we talking about some news article that I've missed?

Anyway, yes, programming has always been in the kids domain for as long as I've been around. But, having been in the IT industry (as a programmer) for quite a long time now, I've seen a lot of changes. One such change is the surprising fall in popularity of computing. Even uni students and graduates are often not good at what they 'want' to do because they couldn't be bothered to study/practice it as a hobby in their own time (I've met a lot of students, too many have a disappointing attitude). So if its on the rise again, then maybe it is fair to say so. Maybe its a political thing. Its a nightmare these days trying to recruit because people either can't be bothered or are no good (or both), because Britain just isn't churning out enough new programmers. Maybe if you tell the youths that all the other youths are doing this, maybe some of them will take an interest and become the next generation of much needed techies.

 

That's all still in the curriculum but under the heading of maths (shape and space). Enormous fun to teach and not very hard, most kids get the basics easily enough but you can use it to really extend their thinking. However, if you only have 3 floor robots and 2 adults per class of 30 it is tricky to deliver and assess.

 

I've been thinking for a while now about a little project for my oldest son (now 5) and me. The 'programming language' (its not really a programming language) that features the robot 'turtle' and commands like FORWARD n is called Logo. Unfortunately it seems to have fallen out of favour as I can't find any decent implementations of it. However its not hard to develop, so I might write my own implementation of it. Which brings me onto the aforesaid project idea. My son shows a natural aptitude for computing and clearly has the technical/logical thought process that seems to run in my family, so I'm thinking when he is a little bit bigger, maybe next year, we're going to build our own robot, fully programmable with a simple implementation of a kind of spin-off of Logo.

I have no idea how schools go about funding stuff, or even if they are allowed any freedom in such matters, but I think loads could be gained if schools did a similar project, at very low cost, for the whole class.

Here's my thinking:

Take one Raspberry Pi (a tiny, super-cheap computer that runs Linux and has all the hardware in place for taking external inputs and outputs as will be needed for a robot's 'brain'). Add some cheap off the shelf electronic components (power transistors to take the low powered signals from the Pi to switch to a higher power needed to drive motors, diodes, resistors, motors and some strip-board - all available cheap from Maplin or RS). Throw in some mechano or similar, and you've got all the hardware.

Then, as a class project, get the kids to work out what exactly they want it to do, then in proper commercial project planning style, work out what is feasible given constraints on time, money and available expertise. Whittle it down to a proper practical plan with realistic milestones. Work out which class members are natural leaders, which ones prefer to be led, who has the aptitude for hardware, who has the aptitude for software, etc etc. Divide them up into teams (a software team, a hardware team etc) and let them get on with it, but of course under proper guidance. With something like that, you not only teach them basic (very basic) electronics, programming, very basic mechanical engineering etc, but in terms of transferable long term benefits, you show them that the stuff they're learning is not just boring theory but is actually practical, they do something interesting which might whet their appetite for further study, they learn team work, and they get an insight into how a proper project works in a commercial setting.

I don't have access to a full class, so my project will be on a smaller scale. Just me and the lad, which I guess will mean I'll be doing the more tricky bits, but hopefully it will still be fun.

 

Lucky lad! It all sounds brilliant, Clue, and after you've trialled it with the boy you might think of telling his school about it. Years 5 and 6 would absolutely love it, and it is exactly the sort of experience they should be having but probably won't. I don't want to be sexist here, but unless they've got one of those eccentric male boffins for a teacher there's probably not a chance in hell. I know how hard it is to get untechnical people to even consider making the most basic sort of model. It's like they're scared of it and they just can't see it, somehow.

So, if you were ever to offer to spend a couple of afternoons in your lad's class, working in tandem with the teacher on the kind of thing you mention above, you might be doing those kids a favour. If you could spare the time, that is. If you put in the science and the IT the teacher could put in place all the organisation.

There, that's your holidays planned for you.

 

That's an excellent idea. I'd be up for that if bureaucracy doesn't get in the way. I'm thinking red tape from having to get the necessary background checks to work with kids etc, possibly insurance burdens etc.

As for getting the time, I reckon I could wangle that.

I'd love to do something like that. So many kids round here at least live up to the 'deprived/underprivileged' stereotype and are therefore a bit naughty from time to time. They have no excuse, but that's another matter. The school does an excellent job of engaging the kids, certainly far better than my school did, but they are obviously limited in what they can do themselves, so if I could lend a hand with a project like this, I reckon it might just work out to be a positive thing for a lot of people.

 

The school might get the CRB check done for you.

 

One thing about the Raspberry Pi is as I'm sure you are aware of, is it's Linux kernel is not a real time one.
For simple I/O using switches, leds and so on, it makes pretty much no difference but as soon as one wants to use anything PWM related (servo motors, tone generation) depending on the frequency and duty cycle of the PWM, the result is that in the case of servos, they will tend to stutter whilst the O.S. goes off and services other things.

One way to overcome this, is to use an Arduino as a slave device.
You could send serial data/bytes from say a Bash script on the Pi to the serial port of the Arduino which in turn generates the PWM at a given rate and duty cycle depending on the data/bytes received.
The Bash script could be called from any app and by using the variable "$1", you can pass any command line parameters to the serial port:

Code:

#!/bin/bash
echo -n $1 >/dev/RequiredSerialPort

So to invoke the script above (assuming it's called OutData.sh and the permissions have been set:

Code:

./OutData.sh 50

Where the 50 is passed to var "$1" and could mean a 50:50 duty cycle for the PWM generated by the Arduino (or what ever syntax you decide for the Arduino code or "sketches" as they are called).

The Arduino is very simple to program (C++) and would be transparent as far as the interfacing goes.
This has the added advantage of hiding the extra layer of complexity from your son whilst he concentrates on the main code/app on the Pi.

 

Ah those were the days when our eldest was making money writing Machine code at the tender age of 11. Long before the BBC machine I might add, Now he designs micro chips for a living. (And a very good living too).

 

Excellent info @DIY-Dave . I was aware of Linux being not quite real time, but in the interests of cost and simplicity, I wasn't going to go for the precision of stepper motors. I was going to go for ordinary DC motors collected out of busted old toys, with cheap plastic gearing.

That would hardly make a super-precise space project but I think it will be quite sufficient for a bit of educational fun.

 

That is a good idea using DC motors, they are widely used and cheap as chips.
In fact it may be worth buying a complete new toy and modifying it for computer control.
I was thinking about those cranes (similar to attached picture) with a wired remote.
It would be quite easy to rip out the switches and replace them with relays (driven via some transistors and snubber diodes).
Kids always like real world things which can be controlled by computer.

 

I hope I'm not the only person only understanding 1 word in 10 of this!

 

My son has two of those. He got one off the car boot sale and someone bought him one one christmas or birthday. Both are doing as all toys ultimately do, lying broken gathering dust somewhere