Something I've been thinking about building for the Mission Planner is some improvements to the battery monitoring during flight. I've been caught out a few times with low battery warnings due to bad estimations of whether I would make it back so I've been thinking about adding something to the Mission Planner to help monitor and estimate the battery levels better. I'm just tossing around what this might look like so would like to hear what others might think or like to see?
Some Ideas I've had:
- A 'range circle': Actually, not my idea. Doug Walmsley is suggesting a range circle displayed on the flight screen - I really like this idea! here is the discussion about it: http://ardupilot.tridgell.net/forum/viewtopic.php?f=12&t=5876 - I would really like Doug involved in this discussion - but this topic was from 2014.
- A count down timer for battery life - time based seems to make some sense, but I think less than distance based. Estimating time is more useful though when you are doing a lot of hovering. It's nice when you're lining up a shot that you can estimate that you've only got 2 more minutes before you have to leave
- A distance count down - more useful when you're not hovering but flying out and back.
- Focus on auto missions and the margin for making it through the mission
- Put some voice feedback of the battery percentage left - this would be a nice feature to help improve battery awareness (for example the 30s voice being able to include battery percentage in the call-out)
With most of these ideas, estimating involves having:
- a good linear and or predictable usage rate (there's no point if the last 20% of voltage discharges in 10% of the flight time for example) - i'm not sure if this is a real problem as I haven't really looked into this, or just requires users to have good quality batteries, or use current monitoring which shouldn't have this problem compared to using voltate estimation
- we need to be good at estimating how much power will be required to get home (for example if the RTL requires you to climb 80m before travelling back into a headwind).
- several fairly accurate variables that need to be configured for a model and an input to account for headwind on the day would be needed
- these variables could even be calculated from a flight log by estimating average current used while hovering vs climbing vs travelling, etc.
- have a screen which just makes it clearer what X% battery actually means - as I'm even a bit confused what this is based on with different models i have (current monitoring vs voltage based) - and what happens when i land and disarm and fly again (i am not sure whether the current remaining resets itself after re-arming - but this is just something i need to learn more about before starting on this project).
I'm thinking that a feature like this wouldn't make sense to be implemented on the APM itself as it would probably be too expensive a calculation to trigger failsafes - and I think (at least initially) the feature would require a learning process to determine how accurate all of this is as we don't want vehicles falling out of the sky due to bad estimates.
What do people think? Does anyone want to get involved with this project - I'm happy to do the programming but probably need some input from more experienced long distance flyers, and fixed wing flyers to make this really work well.
Ben
Replies