Thanks for the detailed reply, Bdale!

   Indeed, I would definitely think twice before touching a flight-tested
firmware, you never know where regressions might creep up with
unexpected/disastrous results... I'll continue thinking about the whole
thing,

Ed

On Sat, Nov 11, 2017 at 7:16 AM, Bdale Garbee <[email protected]> wrote:

> Edouard Lafargue <[email protected]> writes:
>
> >    I understand that once the unit is in Pad mode, it becomes pretty much
> > fully autonomous and won't take commands anymore, which I assume is a
> good
> > security feature.
>
> That was, indeed, the primary motivation for being transmit-only in pad
> mode and during flight.
>
> > That said, there can be situations where it would be
> > useful to revert back to a "safe" or "idle" mode in case launch is
> > aborted.
>
> Yep, we agree.
>
> > Is there any way this can be done on the Telemetrum?
>
> So there are two problems.  I suspect neither is insurmountable, but
> Keith and I haven't ever come up with a good solution.
>
> The first is that once in pad mode, the most important task is detecting
> launch.  Anything that detracts from that is a problem.  Since we're
> using a multi-tasking system, it's entirely possible that we could check
> for a ground command from time to time between telemetry bursts without
> compromising launch detection.  But since our radios are half duplex,
> any time we spend listening for commands is time we can't be sending
> telemetry, etc.  And with increasing complexity comes greater risk,
> which translates into a need for more testing every time we touch
> anything in the firmware to make sure we haven't broken essential
> operation.
>
> The second is that once you've handed control over the LCO, anything that
> might cause the airframe to become *not* ready to launch remotely opens
> up the possibility of an airframe being launched that would not fly
> safely.  There's no way for our boards to know, for example, whether the
> launch system has been put in a safe mode and/or the igniter has been
> pulled such that it's "safe" to go out of waiting-for-launch state.
>
> We have, therefore, just accepted the fact that once you're in pad mode,
> you can't get back to idle using only radio.
>
> If anyone has both strong feelings about this *and* and explicit
> suggestion about how to add the feature in a way that mitigates
> risk... feel free to let us know!
>
> Bdale
>
_______________________________________________
altusmetrum mailing list
[email protected]
http://lists.gag.com/mailman/listinfo/altusmetrum

Reply via email to