Verification

AD_ward9

To kick off the verification thread, here are the present plans/actions:

1. Prove the verification framework. The following link http://www.deeplife.co.uk/files/RB%2..._OR_051106.pdf is a model of a rebreather that is used to run the Verilog and C code that makes the Rebreather tick. I believe this is the most detailed and accurate rebreather model in existence: please advise otherwise, and also of any improvements that can be made.

2. Progressive publication of the FMECA: Vols 1 and 8 are out. Others will be out on roughly weekly intervals to avoid an info flood.

The FMECA includes all test results for compliance testing. These comprise are automated chamber tests, bench tests and backed up by simulation results (that can be more extensive/detailed than test results).

3. A simulator of the whole rebreather, using the model referenced above, driven by the actual code (Verilog HDL and C Code in the microcontroller). The published version will just use the C code, to save people having to load up Verilog tools for co-simulation. The simulator is part of the Rebreather Virtual Test Bench we have, that is being cut down in this manner to make it accessible. Some bugs still in the cut-down version, but a month or two will see these out and you can play with it.

4. Actual dives. When everyone is happy with all of the above. Most subsystems have been dived, though not generally in their present format.
There is a group of people who be lent units from the first production run, to try out on some very extreme dives. Needless to say, we have to be fairly confident before doing that, including having all the automated/chamber test results, and a progressive build up of manned dives.

Cheers,
Alex