Geocaching Project

So, I started out 2/3 months ago using arduino and I was able to build a LED that flashed a pattern. Now, since I have money to start a new project, I would like to make sort of a GPS that would send me a SMS every time the GPS change its location (note that it wouldn't be send messages if the change is minimal). I'd probably add a 3-axis gyroscope to also sense direction and tilt. Can this be build and what parts/ components would I need to use? Note that it wouldn't be connected to a WiFi network and, since it would be used for geocaching, it has to have low power consumption. Probably it would also log the coordinates to a SD card.

How often will you plan on changing the batteries?

All the geocache sites I've visited have been fairly well hidden. If that is the case with yours, how will the GPS see the satellites it needs to see to determine if it has been moved?

Powering a GPS, Arduino, and GSM shield with batteries is going to have you replacing batteries quite often - probably a couple of times a week at least.

Can this be build

Yes.

what parts/ components would I need to use?

The ones you listed - Arduino, GSM shield, GPS shield, gyroscope, accelerometer (why?), batteries, etc.

it has to have low power consumption.

GSM shields and GPSs don't go in the low power consumption column. They go in the other column.

Probably it would also log the coordinates to a SD card.

In addition to sending you an SMS each time it moved more than a certain amount?

Lvothe:
since it would be used for geocaching, it has to have low power consumption

All the features you mention are feasible, but the power requirements would be my first concern. How long do you need it to run for, and what sort of batteries did you envisage using?

GPS is obviously dependent on satellite visibility. Do you need any sort of reliable movement detection, or are you happy to say that movement detection is only possible when it has satellite visibility?

accelerometer (why?)

i don't think it would be required if it used the gyroscope. I needed the accelerometer to spot changes in orientation or direction.

In addition to sending you an SMS each time it moved more than a certain amount?

It goes like this: it would record the location every 1 hour and if the difference between the regular location (predefined) was large, it would send me a SMS. The gyro would record orientation every 30 minutes, and if it saw a large difference between the predefinitions, it would activate the GPS to send me a SMS.

How long do you need it to run for, and what sort of batteries did you envisage using?

At least a 5040 (GPS+ Gyro) plus 5040 (Gyro) seconds per week since i could change anytime I want. The batteries would be small/medium sized ones.

Do you need any sort of reliable movement detection, or are you happy to say that movement detection is only possible when it has satellite visibility?

The GPS can be turned off when satellite visibility isn't good but the gyro needs to be turned on even when the gps is off. Also, if the gyro detects motion, it would send a SMS just saying that the arduino moved and it would save the info every time it turned on.

Also, how would be possible to send the SMS?

Get a GSM shield.

Between a cellular plan and batteries you will win the award for most expensive geocaching prize ever!

The project would be a automated cache.

I probably would use a mechanical tilt-switch.
The device could be build to remain in sleep mode (power saving!) until someone picks it up and the tilt-switch triggers an interrupt to wake it up.

kduin:
I probably would use a mechanical tilt-switch.
The device could be build to remain in sleep mode (power saving!) until someone picks it up and the tilt-switch triggers an interrupt to wake it up.

Exactly. But I also need GPS readings at least twice a day.

An external RTC can wake the Arduino any number of times a day. The Arduino can power the GPS, wait an eternity for the GPS to come to life and return a reading, wake the GSM shield, send the message, and then go back to sleep.

PaulS:
An external RTC can wake the Arduino any number of times a day. The Arduino can power the GPS, wait an eternity for the GPS to come to life and return a reading, wake the GSM shield, send the message, and then go back to sleep.

Yes. But perhaps I'm over thinking, but maybe, instead of using a gps shield, just use the gyro, since I only want to know if the arduino as changed place. And how much measurings (both Gyro and GPS) would a 10gb SD card would be able to save.

And how much measurings (both Gyro and GPS) would a 10gb SD card would be able to save.

How many buckets can my barrel hold? If you want to store one reading a day, you could probably store a couple of centuries worth of readings. Store more often and the duration goes down.

How you store the data matters. How much data you store matters. How often you store the data matters. In which file(s) you store the data matters.

PaulS:

And how much measurings (both Gyro and GPS) would a 10gb SD card would be able to save.

How many buckets can my barrel hold? If you want to store one reading a day, you could probably store a couple of centuries worth of readings. Store more often and the duration goes down.

How you store the data matters. How much data you store matters. How often you store the data matters. In which file(s) you store the data matters.

2 readings a day, .text file, a line where's written time and coordinates.

Lvothe:
2 readings a day, .text file, a line where's written time and coordinates.

At that rate a 10 gigabyte card would last for 20-40 years, assuming each reading uses 64 bytes or so.

kduin:
I probably would use a mechanical tilt-switch.
The device could be build to remain in sleep mode (power saving!) until someone picks it up and the tilt-switch triggers an interrupt to wake it up.

That is what I was thinking. GPS takes power. Arduino/Atmel takes power. GPRS takes power. This can't be running all the time. Shut everything down and run the ATMega on picopower and have it wake up on an external interrupt from some sort of mercury switch plus a few times a day. Then it would be feasible. I would not use Arduino at all, just a bare ATMega or ATTiny "P" chip. All these things take money too, you will be putting something fairly expensive out there into the wild for this project.