Litchi - Minimum altitude & smooth transition between waypoints?

[Mike Toy]

I want to fly a mission based on AGL (15 meters in this case) but it ideally needs to start at at Ground Level – it’s a video project. The idea is to take off, move in a direct line (gradually ascending while moving forward) to WP 2 and then continue to subsequent waypoints all of which are 15 Meters AGL. There are a couple of issues that I can see:

The vehicle will rise to the altitude of WP 2 first and then and then move to it so you get a vertical track followed by a horizontal one - no smooth transition there.

Is there a way to make the vehicle move from WP 1 to WP 2 in a straight line?

If WP 1 is set to Ground Level and I was to load and start the mission from a point 10 meters away at Ground Level (it would be hard to find the exact location of WP 1 if I’ve setup it up in Mission Hub) then what would happen? Normally the vehicle would take off, move to WP 1 and start the mission. But in this case it’s ‘taking off’ and moving to WP 1 which is also at Ground Level – theoretically it would just move sideways at Ground Level to WP 1 which of course would not be good…..or it might take off, ascend to a minimum height, move to WP and descend.....not good either.

Another issue is the accuracy of Google – if it’s out by much the vehicle could theoretically try to descend below Ground Level….

Can you set WP 1 to Ground Level ?

Is there a minimum altitude for a waypoint?

The only way I can see to do it is to set WP 1 at Ground Level (assuming you can do that) and then a series of ‘short’ WP’s each one a meter or so above the prior one and a couple of meters ahead. So in order to ascend in a smooth line to the 15 Meter mark you might need to put in ten different WP’s to create a straight line.

Does anyone have a quicker, easier and less worrisome solution?

Thanks………

 

[RodPad]

Welcome to the Forum!

I'm sure some better tips will come from others.

Do you have a Litchi Account, and been using the Mission Hub?

My first thought is make your second WP at ground level, then cut that part wp1 to wp 2 out of the video.

But I wouldn't ever try to actually get to ground level, your Barometer and GPS are not as accurate as your thinking.

WP at Ground level, negative ground level yes, that's how you run missions that are below your starting elevation.

Rod

 

[Mike Toy]

Yes, I have an account and use Mission Hub. A good idea but if I set WP 2 to Ground Level it may well just fly into the ground......wouldn't really want to try and test it.

 

[RodPad]

I read your first post a few more times.
I have never paid much attention of what happens between start and WP#1.
I load the mission with the motors off.

What I recall,
It takes off straight up I guess about 20' then it heads toward waypoint #! vertical / horizontal, but still pointed same direction as take off, when it gets to Way Point #1, then the direction changes to the setting of the waypoint #1.

Don't know if this helps you, but it did me.

Rod

 

[bsartist]

Regarding moving in a straight line from WP1 to WP2, it will gradually ascend from WP1 altitude to the WP2 altitude while travelling to WP2. It won’t go straight up to WP2 altitude after reaching WP1. In other words it will do what I think you want.

 

[Mike Toy]

Thanks Bill that answers one question but can I set WP1 at Ground Level and start the mission from there? My concern here is that if the vehicle is not positioned exactly at WP1 at take off it will try to fly there - at ground level.....

Is it possible to modify the mission on location, i.e once the Home Point has been registered could I drag WP1 onto it, save it and reload it? Would you need a data connection to do this or will the app be able to update any changes without an internet connection by using the information stored on the iPad?

Maybe best just to ask how you might do it. The mission/video needs to start at take off, that's the key thing. Once it gets to the first waypoint the rest is easy, it's getting it from ground to that first waypoint that is the hard bit. Or am I missing something?

 

[bsartist]

I’m not sure of what your definition of ground level is and some of Rod’s comments are relevant. To wit...

I’d be hesitant to create any waypoints too close to the ground regardless of how accurate google or mission hub is in their heights. I wouldn’t trust the bird to stay within an accurate - or maybe the term here is precise - altitude to keep you from slamming into the ground. I saw a video here of a guy flying low over water pushing the stick forward. He took his eye off of the AC and it descended enough to hit the water. Total loss. The lowest waypoint I’ve created is 8 ft. That was not using AGL. It was just above takeoff point but I was flying from the edge of a lake so pretty close.

I also don’t know if Litchi imposes a minimum altitude. I don’t think I’ve ever seen that though. But it won’t fly up to 20 ft first. That didn’t happen in my 8 ft starting waypoint example. As far as my experience, when you hit Play on the mission it will ascent or descend to WP1 height. Then turn toward WP1 and fly there (heading will be toward WP1). When it reaches WP1, it then swings rapidly to the heading it will fly FROM WP1 on and immediately start flying to waypoint 1.

Let’s say you can accept something like 8 ft. Or If you are willing to try, say something like 4 ft, then a few things come to mind. I don’t have a good idea of what your constraints are in terms of your boundary condition on start of mission. Do you really need to have footage rolling while stopped and then take off and start ascending (as if recording an aircraft rollout)? If so I’m not sure you’ll get it to do that. Even if you use straight turns setting, it might pause at every waypoint. If you don’t need that then Rod’s suggestion is a good one I think whereby you really have another waypoint before where you intend to start your video and just edit out anything before the waypoint that you want your video to start at. In other words, edit out what is now the footage between WP1 and WP2 (old WP1).

Always remember when plotting an AGL mission to plot plenty of waypoints since they provide the resolution for the vertical path. If you want to follow lumpy terrain, you need a fair number of samples (waypoints) otherwise it will fly that straight line right over and possibly into an intermediate hill while on that straight path to the next waypoint. Litchi will only sample AGL wherever you place waypoints. And of course it does not account for obstacles - trees or otherwise.

Finally, naturally this is all just sharing ideas and guidance. It’s obviously up to you to make sure you don’t hit anything with your mission. Too many variables for anyone to give you an assurance that you can do X and achieve Y. At the end of the day, if you are willing to take some risk, then hopefully you can keep it within line of sight - at least during the riskiest parts - and be ready to quickly abort the mission and get it to hover. But if you are trying to plan a couple legs at six inches off the ground then you are unlikely to be able to abort in time to prevent a ground strike which can happen with barometer inaccuracy or a quick downdraft.

Edit: I’m going to hedge on my comment on not first rising to ft. My 8ft starting mission was long enough ago that my memory may not serve me correctly so I’m not unequivocal on that.

 

[RodPad]

@bsartist,
Good Stuff!

Rod

 

[Nightwolf]

Yes, good stuff. Thanks for the post. I'm new to Litchi but i was wondering could OP start in manual, take off manually then hit go on Litchi to achieve the result?

 

[Mike Toy]

Bill,

Thanks for taking the time to send such a detailed reply, that's very helpful. I'm going to have a play around this afternoon and see what I can come up with but I will avoid plotting any near ground-level waypoints for now. Not worth the risk. The alternative (which I was trying to avoid) is to take off from a point behind WP1, fly manually to WP1 (or as close as I can get to it) and then start the mission at a safer altitude (15M AGL). I'd then have to edit the footage to get an even transition between the two parts of the flight. Similar to Rod's suggestion but the manual take-off would remove any chance of flying into the ground if there are discrepancies between Litchi's WP elevation settings and the actual elevation. Wouldn't need to be off by very much. Thanks again.

 

[Mike Toy]

That's what I am going to try later today , I'll keep you posted. I guess it'll depend on what happens between switching from manual to starting the mission. We shall see......

 

[bsartist]

...but the manual take-off would remove any chance of flying into the ground if there are discrepancies between Litchi's WP elevation settings and the actual elevation. Wouldn't need to be off by very much. Thanks again.

I think you’re still missing another important source of error independent of Litchi accuracy. Another important source of error mentioned by both Rod and me is the aircraft itself. It does not have altitude precision or accuracy to the inch and this could affect you even in manual flight as I pointed out by the guy who lost his drone in a lake. You still have not shared what you mean by “ground level.” Taken literally that could mean inches. Or maybe a foot. I wouldn’t trust the barometer or VPS (if so equipped) to hold me that close to the ground.

 

[Mike Toy]

OK, fair enough, I see your point. To clarify I mean ground level to be the ground itself, e.g an elevation of zero. Nothing to do with sea level so if you were to walk up a hill you'd always be at ground level no matter how high you went.

I've just plotted something that involves WP's 1-5 being at 1,3,5 & 10M elevations relative to the take-off point with 5M forward distance between them. WP's 6 & onwards are then relative to ground. Since the first five WP elevations are relative take-off in theory the vehicle should fly in a relatively straight line to the desired height. This would I think be safer than using relative to ground.

 

[RodPad]

Any mission on the hub you can share like this.
Mission Hub - Litchi

Even though you can edit mine, you can't mess up the original.

We might save you from a crashing mistake.

Rod

 

[Mike Toy]

Thanks Rod,

I’ve set up a trial mission here, have a look and see what you think:

Vaucluse

Yesterday I observed the following:

At take-off the vehicle will rise to WP1's altitude first and then move horizontally to the WP itself. Initially I had set WP1 to 1M but it actually only rose to about 18” before heading for WP1. Too close for comfort so WP1 is now at 2M.

Even with short distances between the first set of WP's and a gradual increase in height for each it was still a pretty jerky flight.

Had I read the manual more thoroughly I would have seen that you do in fact need a data connection to change the altitude setting on location – if you try to switch WP altitudes from RTH to RTG you'll get a message saying "Internet is required to compute altitudes relative to ground"

Basically all I’m trying to achieve is a smooth flight from take-off to WP 10 after which the altitude remains fairly constant for the remainder of the flight. Any ideas on how to do this would be most appreciated!

MIKE

 

[RodPad]

I tried a couple of things when I got home some low way points very scary, be careful of my ideas.
I confused about this,
"At take-off the vehicle will rise to WP1's altitude first and then move horizontally to the WP itself."
All of missions I have WP#1 set at 100m, I'm starting the mission with props off, it does not go to the height of WP#1 before it moves to the WP#1 location.
Doing my test yesterday, WP#1 being 10' and 40' to the East almost crashed in the ground (My Yard).
When I can, I really going to focus on the actions of what happens before WP#1.
I have actually learned a couple of things so far.

On your mission, are your really testing at 20 mph?
More way points are smoother but not that close together.

@bsartist,
You got something?

@GrammatonxXXxCleric,
You like low level flying.

Rod

 

[bsartist]

I tried a couple of things when I got home some low way points very scary, be careful of my ideas.
I confused about this,
"At take-off the vehicle will rise to WP1's altitude first and then move horizontally to the WP itself."
All of missions I have WP#1 set at 100m, I'm starting the mission with props off, it does not go to the height of WP#1 before it moves to the WP#1 location.
Doing my test yesterday, WP#1 being 10' and 40' to the East almost crashed in the ground (My Yard).
When I can, I really going to focus on the actions of what happens before WP#1.
I have actually learned a couple of things so far.

On your mission, are your really testing at 20 mph?
More way points are smoother but not that close together.

@bsartist,
You got something?

@GrammatonxXXxCleric,
You like low level flying.

Rod

I need to get on the desktop to view this mission and see if there is a reason for all the waypoints (eg terrain). I did see he had curved turns so that eliminated my first theory. A video would be very helpful to see what the jerkiness is. I’ve also never mixed AGL with above takeoff point elevations in the same mission. Doubt it’s related though.

As far as elevating first to WP1 altitude first before heading there, I’m pretty sure that’s what I see. Not sure why you experience something different. I can give it a try maybe this afternoon.

 

[bsartist]

I took a look at the mission at the desktop. This is really pretty flat terrain so I am not sure why you are using AGL altitudes. The highest point is not more than 5 feet about the takeoff point. Also, since it's pretty flat, you don't need all those waypoints between 1 and 8. It looks like you want a straight path both horizontally and vertically. So I would eliminate those points 2 through (including) 7. It should ascend smoothly from 6ft to 29ft. I am attaching a 3D google earth view of your mission as well as one with my proposed change.

Yours:



Mine:

 

[Mike Toy]

Rod,

I place the vehicle on the ground a few meters behind WP1 which has an elevation of 2M. Then I fire it up and press play. It takes off, rises straight up to 2M and then moves forward to WP1 and starts the mission. I'll try starting it further from WP1 and see what happens although I have a feeling it will do the same thing. So you are saying yours will take off and fly 'straight' to WP1, i.e

I'll also try another version and space the WP's out some more.

The first few runs were done at about 20 kph which gave me enough time to toggle over to 'P' and lift it manually if need be. The videos need to be capped at 30 seconds and each misson will cover somewhere between 350 and 700 Meters so I am trying to find a comfortable speed - just for fun I did one at 45 kph yesterday and did not like it one bit! The real missions are pretty much in straight a line 'though, I just do the tests in a circle so I can see what happens at each WP. Ultimately the video will have to be sped up to squeeze it into 30 seconds anyway but I'd like to try and keep it to a minimum.

 

[Mike Toy]

Hi Bill,

I'm using AGL because the real missions (this is just a practice mission at the flying field) will be flown over terrain that has a lot more changes in elevation. Some will end below the start point and some above and ideally the vehicle should remain at a fixed height above the ground once it reaches it's 'cruising' altitude. So just practice really.

Yes, I'm trying to get a straight path horizontally and vertically - maybe the vehicle only ascends vertically to the height of WP1 as at this point it has not actually begun the mission. If it moves smoothly from WP1 to WP2 then what you've proposed should work although there will still be a video 'lag' of sorts at the beginning - it will take-off, ascend to WP1 elevation, fly horizontally to WP1 and begin the mission. But the smaller the lag the better so I will give it a try this afternoon. I'm not sure that what I want is actually going to be achievable in one smooth flow so it may come down to editing in the end.

Thanks,

MIKE