For the robot project I am using a MotorHawk board, which essentially provides motor control over USB.

Unfortunately the code that comes with it isn’t that great. I knew it was Windows only from the start, but when I had a closer look, it comes with a DLL, and sample applications in managed VB and managed C++.

If you’ve ever looked at managed C++ coming from a non-managed C++ background, you will know how ugly it is. I was not even able to get the project to build. So I set out with a couple of hours to get something nicer sorted out so I can hook it all into ROS eventually and get some real automation going on.

Amoeba-1, my robot for Project Origin which I have been working on, has had a major facelift.

Check out the photos over on Google+ as I dremel’d late into Sunday night and gave it a bit of an overhaul…

https://plus.google.com/photos/107164677276760163853/albums/5666781090277872897

We were in the Apple store the other day (the fairly new shiny one in Reading’s Oracle shopping centre, which seems to have the mystical property of being perpetually busy irrespective of whether the rest of the town is rammed or completely quiet) swapping out a busted iPad power brick – they kindly gave me a new one from their spares for free, after I convinced them I am capable of debugging electrical issues and they verified it was busted – when I decided to buy one of the new iPod Nanos.

Nano

I could claim this was an impulse decision, but that wouldn’t be entirely accurate – I’d been tempted by one of the Nanos since it was first released, but thought the software could be improved with more functionality. On 4th October Apple did just that, releasing a software update for the Nanos as well as a price drop (although the hardware was not updated), which was just about enough to tip me over the edge and grab a couple (only one for me!).

This afternoon, our robot project was supposed to have a demo. What actually happened was that we talked about it, switched it all on, and then it never moved. The demo was scrubbed, and fortunately this year did not count for any marks – otherwise we would have ended up getting very low numbers.

So why did a project which had been working so well fail so brutally? It was not a failing of the robot, nor really a failing of any one team member (although the fact I got the configuration wrong and failed to realise in time was the cause of the issue). I’m going to attribute this one to bad planning and last minute work.

I’ve been busy hacking around with Kinect, the Xbox 360 “controllerless controller” I’ve been writing a lot about recently.

Yesterday I got it up and running under Ubuntu, and after that I got a very simple “nearest object” algorithm running on it. Check out the videos below! Also – the code.