Monday, February 25, 2008

Finally some forwards motion

The GPS receiver arrived today and thus ensued many hours of unproductive fun until I decided to RTFM. It turns out that it works incredibly well, he said as he stared down at himself on Google Maps. Unfortunately the receiver doesn't appear to have the Ion Cannon functionality installed.

Once the excitement of the receiver had died down I started coding and now have a fully functioning home-baked GIS system. It calculates a bounding circle and searches for bookmarks in a ~60m radius around the user and returns the most likely location.

In the event that there are numerous possible locations I intend to add vector calculations to find which bookmark the user is most likely close to but that is for another day.

Thursday, February 21, 2008

Round in Circles

The GPS device has been ordered from the mighty PC World.

I've been looking through a number of GIS solutions and had settled on a Java API for Google Earth which was going well until they told that it didn't work. A whole days work down the pan quicker than you can say .Net might not be that bad.

So now it's back to look for different GIS systems. Who knows, I might even find one that works.

Saturday, February 09, 2008

File Appropriation

Since abandoning the mobile phone idea I've had some spare time to get the fundamentals of data reporting into place. Using pseudo-random numbers and timestamps the LittleBrotherClient can now report with a time and a location. Obviously, once I've got my head around the GPS Receiver issue it will be able to report where it actually is.

LittleBrotherServer can also look up where these reports are stored. Once it has got the list of new files from the repository it can open and parse them, so we now have some data to play with on the Server side. When it's finished parsing it can remove the files from the store.

Mobile Programming

In all fairness, this post should have been made last night but due to annoyance it wasn't. I spent Friday afternoon and evening coding prototype mobile applications with the hope that I'd be able to use my Nokia N95's built in GPS Receiver. Things were going well until I attempted to actually use the GPS...

It turns out that mobile programming is quite straight forward once you get round the annoyance of having to display everything graphically. System.out.println simply won't cut it here, we need graphics canvases and the like.

The documentation that comes with the phone isn't that great and alongside a tiny array of programming examples it would be impossible to fit the required development into the time afforded by the University so this idea has quickly found it's way into the Recycle Bin.

In other news, the project plans were handed in today.

Tuesday, February 05, 2008

Episode 1

The topics have been finalised and I'm officially now creating stories based on GPS co-ordinates. In addition to this Ross replied and I now have some direction with regard to programming GPS systems using Java.

The project plan is done - probably with minimal (probably quite a lot if I were to be honest) amendments to be made once Ehud, my supervisor, has had a read through.

I've also had a look at programming using J2ME to see if I can attempt to use the GPS built into my Nokia N95. So far my attempts have resulted in a PNG file depicting a ball bouncing around the screen. Although it only runs inside the J2ME emulator... not on the phone... I assume all shall be well and the department will invest in some GPS hardware for me.

Thursday, January 31, 2008

What's in a name?

Today was a fairly uneventful day so far as the project. I contacted my supervisor regarding the kind of resources that we have available and he put me in touch with Ross Turner who is a PhD student at the University. Seemingly he's done a lot of work with GPS including a project called `Roadsafe`, so hopefully he has a lot of information that I can put to good use. I'm still waiting to hear back from him.

Software wise the Eclipse Java IDE has been installed and configured to make use of Tortoise SVN for version control. This will make things easier when I realise that I'm marching swiftly down a blind alley and need to retrace my steps in order to proceed in a more `correct` fashion.

The project is to be called `Little Brother` in a nod to George Orwell's 1984. I find the whole tracking people through GPS a little scary - particularly as we're writing stories about the things they get up to...

In the beginning...

One of our lecturers has suggested that we write a blog to record our thoughts and activities during our final year project. Doing it this way will mean that I don't have to continually fill in forms detailing what I've done. I'll also have someone to blame if there's a disk failure somewhere and all my work goes to hell.

So, a bit of background on the project. This week has been mainly filled with meetings while we try and decide who is going to do what and what is to be done. The foundations of the project was always going to be GPS and analysing the data provided by it; the main task being to decide what to do with it once it had been analysed. The two ideas were to either relate the information to a map or to produce a story based on where the tracking device had been during the day (given that I'll probably be tracking students the majority of entries will probably relate to The Bobbin - a well loved pub just off of campus). At the time of writing it is likely that I will be using some Natural Language Generation to compose the stories which can later be edited to a limited extent by the end user.

The project ideas have to be finalised (....or our supervisor would like them finalised by Monday) so I guess that's when the bulk of the work will begin. Now then... where do I start...