Torro Conference and moving forward…

Well some of you may know that I volunteered to stand up and talk about Touchdown, the Pod and SOA this afternoon. I never feel comfortable with this sort of thing, generally engineers are best kept as far awa from the public as is possible, putting one in front of a room of people really seems like a bad idea.

No one died, the message went over, I’m pretty sure it was well received and no one has arrested me…yet.

So the interesting thing here is that theres been a development. SOA was always intended to have UK and US versions. the UK version wasnt a priority, turns out this may have been a mistake.

Two messages seemed to come through today. One, that data is being lost. Too much reliance is being put on people’s memories, bits of paper and anecdotes. A large proportion of what Touchdown proposes centres around data retreival, integrity and storage and automating all of this as much as possible.

The other is that there is an apetite for something Like SOA here in the UK. With the move from STM32 to the Raspberry Pi the proposition that you could build your own and just use the Touchdown software is quite feasable. In fact we’ve had a large response to this idea.

SOA as it stands does provide a framework for doing a lot of this and with the move to Pi a lot of the special hardware isnt needed, making it modular is perfectly feasable and with the Pi, if you can use jump wires then its easy enought o get something running.

So where are we…

As the system stands a lot of the applications still need to be ported over, we’ve re-written a lot for Pi already so theres nothing horrid there, it can be done and its not a difficult job.
SOA was never designed to be ‘hackable’ so we need to make things as user friendly as possible, I’m thinking ‘appliance’ here so it can be plugged in and mucked with if needed. Although the Pi can run a monitor theres no need to actually do it if theres another computer to hand.

Sensor wise, its down to the end user. There are a set family of devices we use but adding support for others shouldnt be hard. Keeping it same and sensible is important so we’ll likeley pick just a few limited ones.

Reporting should be automated. With the user getting warnings and them being reported back to a server. Maybe an expansion on the reporting system planned for SOA so people can instantly log weather conditions that are out of the ordinary.

If its possible to maintaie a stable, syncronised time source through GPS or NTP then the possibility of lightning ranging and triangulation is opened up. It also allows for storm trends to be analysed as they happen, a potential powerful tool for tracking major storm systems.

On the back of it I think we will look into building a reporting system to use with this along with a repository for reporting severe weather events and ways of automating certain reports and alerts. On top of this a simple to use website to allow everyone access to the data.

Looking forward the easiest way would be to offer either the pre-built units or instructions, or even both. As all Pi’s are essentially the same a single SD card image can be used with all the necesary software, setup and modifications already done.

Its a very interesting proposition and one I think will be getting looked into over the next few months.