Hey all,
I've been asked to build a rain alarm for my mum. She's partially sighted and whilst she does cope very well, sometimes she doesn't quite see the rain in time to bring her washing in which usually results in it having to be partially dried again.
I know absolutely nothing about the weather, but I've come up with 2 ways of doing this that I wanted to get a more informed opinion on:
1: Purchase some form of rain sensor, find a way to connect it to a computer and write a program that is run when the sensor is activated.
Prose:
Should be very accurate since it would be placed in our garden
Prices for sensors seem to be around £26, making it much cheaper than buying a complete station
Size is a bit of an issue in the planned location and the sensors seem quite small
Cons:
It may require a bit more hardware hacking than I've done before; most of the sensors trigger a relay and I have no idea about cheap ways of monitoring relays over USB
The program will need to be able to distinguish between a single drop of rain and more worrying amounts. This *shouldn't* be that much of a problem to code, but it will require a large amount of real life testing before it works well
A small amount of drilling will need to be done to root wires through the window frame since wireless sensors are noticeably more expensive
2: Find a source for weather a few miles away and have the program look at this information to determine if the alarm should be triggered or not.
Prose:
No investment
The program could be run on my server thereby negating the need for a computer to be on 24/7
It should be able to sound the alarm just before it starts raining, the idea being that mum will be able to rescue the washing before it starts
Cons:
Potentially inaccurate
My questions:
How reliable would the second method be - e.g. if I found a station near to me and the station reported that it was raining, how sure could I be that the rain will travel in my direction? Is there any vaguely accurate mathematical way of determining this? Now that I've written it down it sounds quite stupid, but this method would be much easier to implement.
Has anyone done anything like this before? Does anyone have any tips on monitoring relays over USB or anything else to do with the project?
I plan to open source all of the code once it’s done in the hope that other people will be able to benefit from my work, but as I said above, I wouldn't mind a bit of knowledgeable input before I start buying things.
Any pointers at all would be absolutely amazing.
Cheers,
Ben.
