This 3G modem has caught my attention as a real possibility to become a "component" replacement for the "XBee" and "3DR" class telemetry radios most of us are using.
It comes from Cooking-Hacks.com, you can see the catalog page here -
The support page is here -
The manufacturer's page for the SIM5218 modem is here -
There are three things that make this modem attractive. First, it is a 3G platform and is significantly faster than the predecessor SIM900. Second, it has a complete TCP/IP stack and many important/useful application protocols are also built-in. Third and most important, it has a built-in script processor that runs "LUA".
A lot of time and effort has gone into getting the APM code modified so that it will establish and maintain an internet connection. The problem is that this effort here (Telemetry over IP) is not part of the main APM project. As a result, it is always going to be release dependent and the APM code can be changed in the future in a way that makes all this development incompatible.
The SIM5218 has a scripting language that is included for just the kind of project we are proposing here. The Cook-Hacks shield has an SD card socket and that's where the script code would go. Whenever the SIM5218 starts up, it will execute a named file on the SD card. All of the logic to create and maintain a connection to the Mission Planner could go into LUA script code residing on the SD card. This approach would create a separate modem "component" that would manage the internet connection without modifying the APM code.
The SIM5218 also supports live video. The camera looks pretty simple, but it could provide a low speed backup/extension to the traditional FPV link. Again, it's over the internet so the physical distance between the plane and "pilot" can be greater than we usually consider.
The biggest negative is the modem's cost. Cooking-Hacks sell it for 149 Euro and their US distributor sells it for $199. This more than a new APM kit costs. It would only be cost competitive if the video link would work.
Does anyone else here have any comments on this approach to the telemetry-over-ip project?