History

This project is not new. I have been working on it since about January 2009.

WHAT IS IT?:
Put simply, microLOG is a data logger. This means it measures a sensor in relation to time. Presently I have only used a room temperature sensor, and the data logger would log the temperature every 20 minutes or so. It stored the data as a *.csv file, and this could be imported into excel for graphing.

ORIGINAL IDEA:
A system to log data, and transfer it to a Casio fx-9750G PLUS - a graphical calculator used in New Zealand schools.

CURRENT IDEA:
A system to log data via wireless nodes, create a *.csv file, and upload to twitter. The system can be configured by a web interface.

WHY THE CHANGE?:
The communication protocol for the Casio calculator was too much for my PICAXE, and too much for the mighty Basic Stamp 2. The Casio expects a verification header, followed by a data packet, and then an end header. These data packets all had checksums which luckily were just sums. The checksum had to be generated, the data sent, and only then would the calculator accept the data. I never managed to get this to work...

SOLUTION:
I ended up with a VDRIVE2 - A USB flash drive to SPI / serial 'converter'. I was able to get a PICAXE to append data to a *.csv file on the flash drive.

I saw mightyohm's tutorial on how to hack a wireless router into a WiFi Radio. I also saw Andrew's photoset on using an OpenMESH router as a way of uploading data from a PICAXE to the internet. So I combined the two and ended up with a router running OpenWRT, running a python script, running an http server, and reading data coming off the HopeRF.

HOW MUCH TIME WAS SPENT?:
This is the major project of mine, so it takes up the majority of my free time. I spend a good few hours a week on it, and it has been over half a year SO: 4 hours a week * 40 weeks = 160 hours!

HOW MUCH MONEY WAS SPENT?:
Tough one, as I don't write it all down... I would say about $50 worth of PICAXE gear. A $12 LCD screen, $150 worth of wireless units, $60 for the router, $40 for the VDRIVE, and maybe $20 worth of misc stuff. That lots totals $332. In the competitions I have entered, I have won $485. So I have made $153. It's awesome to think that I have made money on a project I built in my spare time. There are more competitions to enter it into, and this could win more cash!

As a side note, I do this for the project, not for the money. Much of the money ends up invested into the project, or into the hobby. The major purchases with the prize money are a $220 Digital Camera, $75 worth of more HopeRF units, $60 for the wireless router...

WHAT AM I WORKING ON NOW?:

• I'm still building the web interface.
• I am upgrading from PICAXE to AVR - bloody ATtiny series doesn't have USART (only in ATtiny2313 which doesn't have ADC). Lucky for me, the ATmega48 costs a mere $2AUD and comes in a TQFN package. I can develop on a breadboard with my PDIP ATmega48, and then etch a board, and solder on the TQFN. SMD soldering is actually pretty easy!
• I'm working on adding more sensors. I have a windmill, so I will measure RPM, voltage, and current (shunt resistor and ADC I guess..). I can also wire some push buttons to the garage, and log the opening and closing of the garage doors.
• 'Reverse Tweeting'. You can txt a direct message to microLOG, and it will perform a task (maybe I'll log the messages too!)
  

APPLICATIONS:
I tell lots of people about his invention, and many say "It can be used for security". PIR sensors can be wired in, and a tweet can be send when one of these sensors measure activity.

Personally I think it can be used in schools. Students will be able to do science homework (Aren't I nice!?) during the holidays by following the humidity and temperature of their plant. A sensor node will cost about $43 (ATmega48 $2.50, HopeRF $35, PCB $2, Temperature Bead $0.50, Misc Parts $3). Each node would be able to monitor multiple sensors (8 ADC channels). If a student were to monitor humidity and temperature, then 4 students could share a node.

$43 may sound like a lot, but I talked to my science teacher, and our school spent a good few $1000 on their data logger. It is single channel only. My system has the benefit of 8 analogue channels per node, and (in theory) and infinite number of nodes. I say in theory, as wireless means I can have as many nodes as I want. I think I would have a problem with 'collisions'. More than one node talking at a time. There could easily be a work around... Disable the wireless unit of the nodes when there is data being received.



More posts to come.