Originally Posted by DominickC
@EricH - As it stands, the IFI system is perfect for running nearly any robotics project that you might want to do. Granted the serial interface is a pain, it's still a fully functional and robust microprocessor. While it won't do what we ask of our cRIO's today, it's earned my respect.
What about the use of an Arduino or similar type microprocessor?
Arduinos are great for what they are: a tool to teach people about electronics, instrument some sensors, make lights blink, and multitudes of other cool things but they are severely limited in both speed and memory to perform advanced tasks. I don't have the datasheets in front of me but I doubt that an Arduino is as powerful as the old IFI system.
That said there are multitudes of microcontrollers (and similar devices) out there that are plenty powerful (ARM, PIC, ATMEL/AVR, Freescale, pick an architecture!). What they don't get you is the "plug and play" features that we get with the cRio and associated tool chains that effectively help level the playing field. Instead of figuring out the low level implementation of say vision processing, we get tool chains that "work" out of the box. Sure there is a lot of fine tuning involved to perfect these out of the box utilities but the implementation is there for any team rookie to veteran to use. That is a huge advantage of the current system.
Where I think we are running in to problems as someone mentioned earlier is that no one understands the system top to bottom, controller to FMS and all the components between like they did in the IFI days. Sure, individual subsystems are well understood and characterized but do they all play nicely together? Is the cRio IP stack (implementation) implementing something that isn't compatible with the D-Link? The D-Link to the cRio? The D-Link to the FMS access point? This is the area that seems a bit grey to me and may be contributing to the unexplained issues. 95%+ of issues are easily explained (loose power connector, bad wiring, loose connection etc.), it is the others that are puzzling and I'm not sure have a known root cause. I'm not saying anything unexplained caused the issues on Einstein, just that there appears to be an underlying issue somewhere that can't be pinned down.