> I'm working on a project to take a number of environmental and aquatic
> parameters. I was asked to design/spec a system that takes remote pictures
> and uploads them. They had their game cameras stolen, and I think are
> looking for something of an instantaneous backup over cellular 2g/3g/4g for
> the automatic picture captures. I'm aware that game cameras with this
> capability built in are available, but I was informed that they wouldn't
> work for this project - too low of quality. I'm told that resolutions in
> the 3-4 megapixel range are sufficient, and it should just need to be
> captured on a timer. Remote capture is a bonus, but not necessary. Our
> monthly upload should be below data caps. I'm wondering if any of you have
> suggestions of hardware to use.

> Generally speaking, the budget can cover reasonable costs (a few hundred
> for one-time hardware purchases is fine, and a data plan is a given). I'm
> open to hardware in the open source and proprietary/off the shelf realms,
> but my electronic engineering is pretty weak, so anything more complicated
> than minor sautering is probably out since I don't want to fry hundreds of
> dollars of work equipment. The location has no power hookups or other
> infrastructure, and it will need to be able to stay out for at least a few
> months at a time, including in the rain, with no human intervention.

> I have two current plans to do more research into:

>    1. Android based phone with data plan, autocapture application, and
>    solar panel to charge constantly over USB. I think I can find one where the
>    camera quality is sufficient. I'd need to build an enclosure to
>    weatherproof it.
>    2. Lower end canon camera with an Eye-Fi card and CHDK installed to
>    autocapture. The Eyefi card would transmit to a 3g hotspot created by
>    something like a Verizon MiFi or a Sprint Overdrive. Both would need solar
>    chargers, and I'm skeptical if that can keep the equipment on if it needs
>    to support a wireless network. I also think battery charging could become a
>    problem for the camera. This system has more complexity and points of
>    failure than option 1, but gives more camera power.
>    3. Suck it up and use the commercial game cameras as an integrated
>    solution and add better security to prevent theft. What's important is that
>    the loss of a camera does not also include data loss - we can replace the
>    cameras.

> So, those are the options I haven't ruled out ( I ruled out some that
> would require far more tinkering/sautering/complexity that I'm probably not
> capable of). If I've left out any important details, I'm happy to fill in
> any gaps. If any of you have other ideas that I might have missed, I'd
> appreciate pointers in the right direction. Thanks for your help!
> Nick Santos
> Junior Specialist, Center for Watershed Sciences
> nrsan...@ucdavis.edu <n...@enviroconsumer.org>