ACCU Conference 2008, Day 1
So it’s April, I’m in Oxford… it must be the ACCU conference again. This is a leading conference for programmers in the UK and Europe and brings some of the brightest and best in one place for four days of talks, meetings and after-hours wining and dining. I got off to an inauspicious start by spraining my ankle before arriving and I’ve spent the conference hobbling around. Unfortunately, many of the talks are upstairs, slowing me down even more.
The first days talks cover a variety of subjects, covering network services and programming, agile development, programming methodology and robotics.
For the morning session I went to Roger Orr’s talk on “Programming in a Networked World”, where he showed us some of the pitfalls of writing programs that had to function over network connections. The talk was quite high level and more an overview of the possible problems of latency and bandwidth issues, but unfortunately didn’t have time to cover solutions except in a general sense.
I chose the afternoon track on robots, something that sounded fascinating and unusual. Bernhard Merkle started with an introduction to his work at Sick AG, a German company that makes sensors, essential to any robot. In fact, on the control side, a robot is a collection of sensors (input devices) and motors (output devices) and the coordination between them. After a quick look at what his company has been doing as part of the DARPA Urban Challenge, he then took us into the depths of Microsoft’s Robotics Studio, a complete development environment and simulator for creating robotic devices. It’s big and it’s complex, and it’s probably very useful for a large enterprise development, but it seems overkill for someone just starting up with robots.
Which led nicely into the second robotics session, from Ed Sykes and Jan-Klaas Kollhof. They had brought in a couple of Lego Mindstorms robots and we watched them motor around the floor while Ed and Jan explained how they were programmed to do what they did. They had each taken a different approach to the programming, Ed using MS Robotics Studio, and Jan using an open source environment for similar ends. While MSRS allowed Ed to create and control his robots through a graphical interface and then simulation testing, Jan took the approach of writing external scripts to control the robot. Firstly through using Python scripts running on his laptop and communicating via bluetooth and next using NXC and pblua to compile and push the code on to the robot, making it run autonomously. Lego robots provide a very open and flexible environment for playing around with this sort of thing and it was impressive how little coding it took to have them following a line on the floor. With addition of a camera to one of them, it could then follow around the other robot (which had an orange ball on top as a identifier).
For the evening, we gathered in Joe’s Bar and Grill in Somertown, where I can recommend the veggie burger. The others seemed happy with their meaty variations.