Question for the Mission Planner/Tower gurus

[User Name]

I've searched the web but haven't found a definitive answer to this. Say I want to plan a one way flight, taking off at one location and landing at another. Easy enough, just add a "Land" waypoint at the ending location.

But how do you handle if the landing location is at a significantly higher elevation than the take off location? I just tested this with my little 250 quad, and it doesn't switch to the slower, final descent speed until it gets 10 meters (or whatever) above the takeoff altitude. What that means is if the landing location is more than 10 meters higher than the takeoff elevation, it hits the ground at the faster descent speed. Not a huge issue on my little quad, but I'm not sure I'd want the Solo to land that hard with the gimbal and all.

I've read mentions about an altitude offset somewhere, but can't find any specific information about where it is or how to use it.

There's a nice windy river nearby that I'd love to fly down, but it would definitely require a flight plan as opposed to manual control with all the bends, and the ideal starting and ending locations are at different elevations.

 

[Adimuzio513]

Interesting question, can't say I know an answer, but I'll stay tuned to find out if someone else does.

 

[Jubalr]

I've searched the web but haven't found a definitive answer to this. Say I want to plan a one way flight, taking off at one location and landing at another. Easy enough, just add a "Land" waypoint at the ending location.

But how do you handle if the landing location is at a significantly higher elevation than the take off location? I just tested this with my little 250 quad, and it doesn't switch to the slower, final descent speed until it gets 10 meters (or whatever) above the takeoff altitude. What that means is if the landing location is more than 10 meters higher than the takeoff elevation, it hits the ground at the faster descent speed. Not a huge issue on my little quad, but I'm not sure I'd want the Solo to land that hard with the gimbal and all.

I've read mentions about an altitude offset somewhere, but can't find any specific information about where it is or how to use it.

There's a nice windy river nearby that I'd love to fly down, but it would definitely require a flight plan as opposed to manual control with all the bends, and the ideal starting and ending locations are at different elevations.

The Solo would be using the barometer, so landing would not be a problem. It wouldn't matter if the LZ was higher or lower. I haven't tried or had the need for a 1 way mission yet, but may try it just for fun.

 

[User Name]

The Solo would be using the barometer, so landing would not be a problem. It wouldn't matter if the LZ was higher or lower. I haven't tried or had the need for a 1 way mission yet, but may try it just for fun.

I should clarify the issue, especially if you'll be testing it! So here's the scenario:

Take off elevation is 100 meters above sea level, Solo records this as "zero". If I return to this location and auto-land, Solo descends quickly to 10 meters (or whatever the set altitude is) above "zero", which would be 110 meters actual, then descends this last 10 meters at the slower final landing speed.

Now, let's say I'm taking off at the same location but landing at another that happens to be 120 meters above sea level. Solo records the take off location as "zero" elevation, flies my flight plan, climbing along the way. The last waypoint brings it to 150 meters altitude and then it's set to land. It's 150 meters above the "zero" takeoff elevation, but only 130 meters above it's present location. It'll descent quickly until it reaches 10 meters above it's set "zero", which would be 10 meters above it's take off location - 110 meters actual, but this would be 10 meters below it's current location's elevation of 120 meters actual, and it would reach the ground before ever switching to the slower descent speed.

I don't see how the barometer would be able to determine altitude above an arbitrary elevation (pressure 10 meters above a 120 foot elevation is not the same as 10 meters above a 100 foot elevation).

There has to be a way in the flight plan to specify that the landing location is X meters elevation, so it knows to slow to the final descent speed 10 meters above that elevation, not above it's starting elevation.

I suppose we could hard code the slower final landing speed to be 10 meters above the difference in takeoff/landing elevation. So if we want it to transition into the slower speed at 10 meters, and there's a 20 meter increase in elevation between takeoff and landing locations, set the slower landing speed to be 30 meters. If, for some reason it had to return to launch instead of completing the flight, it would just descend slowly from 30 meters. I just think there has to be a better way to handle this.

 

[Jubalr]

I should clarify the issue, especially if you'll be testing it! So here's the scenario:

Take off elevation is 100 meters above sea level, Solo records this as "zero". If I return to this location and auto-land, Solo descends quickly to 10 meters (or whatever the set altitude is) above "zero", which would be 110 meters actual, then descends this last 10 meters at the slower final landing speed.

Now, let's say I'm taking off at the same location but landing at another that happens to be 120 meters above sea level. Solo records the take off location as "zero" elevation, flies my flight plan, climbing along the way. The last waypoint brings it to 150 meters altitude and then it's set to land. It's 150 meters above the "zero" takeoff elevation, but only 130 meters above it's present location. It'll descent quickly until it reaches 10 meters above it's set "zero", which would be 10 meters above it's take off location - 110 meters actual, but this would be 10 meters below it's current location's elevation of 120 meters actual, and it would reach the ground before ever switching to the slower descent speed.

I don't see how the barometer would be able to determine altitude above an arbitrary elevation (pressure 10 meters above a 120 foot elevation is not the same as 10 meters above a 100 foot elevation).

There has to be a way in the flight plan to specify that the landing location is X meters elevation, so it knows to slow to the final descent speed 10 meters above that elevation, not above it's starting elevation.

I suppose we could hard code the slower final landing speed to be 10 meters above the difference in takeoff/landing elevation. So if we want it to transition into the slower speed at 10 meters, and there's a 20 meter increase in elevation between takeoff and landing locations, set the slower landing speed to be 30 meters. If, for some reason it had to return to launch instead of completing the flight, it would just descend slowly from 30 meters. I just think there has to be a better way to handle this.

Actually, you are cutting the barometer a little short of it's abilities. Very easy to see how accurate it is; put your Solo in manual. It is now using the barometer to maintain altitude. Is it varying wildly by feet? Nope, it stays where you put it. Now raise the hover by just a foot or two, and it stays right there at the new altitude. Solo does not use GPS to determine altitude for landing. And it is a good thing it doesn't, altitude is also the least accurate of GPS coordinates:
"Altitude error is specified to be 1.5 x Horizontal error specification. This means that the user of standard consumer GPS receivers should consider +/-23meters (75ft) with a DOP of 1 for 95% confidence. Altitude error is always considerably worse than the horizontal (position error)."

In relation to your scenario above, Solo will not 'climb' to the next WP unless you tell it to in the WP. All WPs are AGL above home point. It does not do terrain following. If you take off to 95' and fly a mission where one of the WPs is 100' above your home point, the earth will reach up and smack thee!

 

[User Name]

Actually, you are cutting the barometer a little short of it's abilities.

No, I'm quite amazed at how accurate the barometer is, but it's not applicable to the scenario I'm describing.

Here's a different example to illustrate - let's say we take off, climb to 50 meters and stay in a mode that maintains altitude (AltHold, Loiter, etc.). If we fly straight towards a 100 meter tall hill... at some point, we'll fly right into the side of the hill at whatever the intersection is between the hill's elevation and 50 meters above the takeoff location - the Solo will not climb over the hill maintaining a 50 meter altitude over the ground at all points.

Atmospheric pressure, which the barometer senses, is variable by altitude relative to a fixed location - not by altitude over an arbitrary location. In other words, and this is an extreme example, atmospheric pressure 10 meters above a Caribbean beach is not the same as 10 meters above the summit of Mt. Everest.

If we take off at one location, Solo records that as "zero". If we fly up a hill, Solo's barometer senses an increase in altitude and since, as you say, it only uses the barometer to determine altitude, it has no way of knowing what the altitude is over the top ground level of that hill, it only knows it's altitude over the take off location at the bottom of the hill.

 

[KRVer]

No, I'm quite amazed at how accurate the barometer is, but it's not applicable to the scenario I'm describing.

Here's a different example to illustrate - let's say we take off, climb to 50 meters and stay in a mode that maintains altitude (AltHold, Loiter, etc.). If we fly straight towards a 100 meter tall hill... at some point, we'll fly right into the side of the hill at whatever the intersection is between the hill's elevation and 50 meters above the takeoff location - the Solo will not climb over the hill maintaining a 50 meter altitude over the ground at all points.

Atmospheric pressure, which the barometer senses, is variable by altitude relative to a fixed location - not by altitude over an arbitrary location. In other words, and this is an extreme example, atmospheric pressure 10 meters above a Caribbean beach is not the same as 10 meters above the summit of Mt. Everest.

If we take off at one location, Solo records that as "zero". If we fly up a hill, Solo's barometer senses an increase in altitude and since, as you say, it only uses the barometer to determine altitude, it has no way of knowing what the altitude is over the top ground level of that hill, it only knows it's altitude over the take off location at the bottom of the hill.

Thinking - not yet tested: Consider doing a change speed "waypoint" before your final landing. I'll test this today - have a hill nearby that I can try landing on with Tower - will be prepared to override if Solo is coming down too hard and fast. I'll report back.

 

[Jubalr]

No, I'm quite amazed at how accurate the barometer is, but it's not applicable to the scenario I'm describing.

Here's a different example to illustrate - let's say we take off, climb to 50 meters and stay in a mode that maintains altitude (AltHold, Loiter, etc.). If we fly straight towards a 100 meter tall hill... at some point, we'll fly right into the side of the hill at whatever the intersection is between the hill's elevation and 50 meters above the takeoff location - the Solo will not climb over the hill maintaining a 50 meter altitude over the ground at all points.

Atmospheric pressure, which the barometer senses, is variable by altitude relative to a fixed location - not by altitude over an arbitrary location. In other words, and this is an extreme example, atmospheric pressure 10 meters above a Caribbean beach is not the same as 10 meters above the summit of Mt. Everest.

If we take off at one location, Solo records that as "zero". If we fly up a hill, Solo's barometer senses an increase in altitude and since, as you say, it only uses the barometer to determine altitude, it has no way of knowing what the altitude is over the top ground level of that hill, it only knows it's altitude over the take off location at the bottom of the hill.

OK, I misunderstood your original point you wrote. As stated, GPS is not used only barometer for altitude. So if flying to a point that you designate as a landing point and the height is higher than home point by more than 10-15', then I would change the speed of landing parameter in tower. For example, I fly from home of 0 altitude and set a waypoint on a hill that is 100' higher and mark it as a landing point. I could have it climb to 130' on the way there and set the descent speed for landing (LAND_SPEED) to a speed that would allow for a soft landing. When Solo landed and detected it was no longer descending, it would then shut down.

EDIT: Sorry, KRVer. I didn't see your post before I posted.

 

[User Name]

For example, I fly from home of 0 altitude and set a waypoint on a hill that is 100' higher and mark it as a landing point. I could have it climb to 130' on the way there and set the descent speed for landing (LAND_SPEED) to a speed that would allow for a soft landing. When Solo landed and detected it was no longer descending, it would then shut down

Ok, so that would basically just designate the slower final landing speed for the entire landing descent - it would come down at the slower speed from the beginning of the descent, not just the last few meters.

I was just thinking there might be a way to specify the elevation of the landing site and have it automatically adjust the final descent altitude to match.

 

[Jeff Vaughan]

The barometer on the solo is really good but you will probably need some attached to the solo that will tell solo what AGL (above ground level) altitude it is flying at in order to maintain a set terrain following altitude. This is something we use in commercial and military aviation in order to determine our altitude above the runway during landings. With the open operating system maybe someone could develop something to attach to the accessory bay. Just a thought.

 

[User Name]

The barometer on the solo is really good but you will probably need some attached to the solo that will tell solo what AGL (above ground level) altitude it is flying at in order to maintain a set terrain following altitude. This is something we use in commercial and military aviation in order to determine our altitude above the runway during landings. With the open operating system maybe someone could develop something to attach to the accessory bay. Just a thought.

The Pixhawk has the ability to download terrain data and reference it during autonomous flights, but I'm not exactly sure how to verify that it's being used and, of course, it's only as good as the data being used. When I upload a mission to my Pixhawk quad, I can remove the SD card and find a file in the terrain folder, so there's something there. Alternatively, the Pixhawk can make use of sonar sensors to detect close proximity to the ground.

But how much of this applies to the Solo, I'm just not sure. I'll have to experiment with my little quad. My initial test last night revealed that it indeed doesn't by default take into consideration the elevation under a landing point.

 

[DavidFisher]

Nothing helpful to add. But mission planner does have a terrain comp mode. But as jubalr mentioned on my post yesterday, its not very accurate as it relies on the map data. I would be very interested in an accessory bay terrain comp sonar thing.

 

[User Name]

Nothing helpful to add. But mission planner does have a terrain comp mode. But as jubalr mentioned on my post yesterday, its not very accurate as it relies on the map data. I would be very interested in an accessory bay terrain comp sonar thing.

I've been doing more research on this. Terrain following, as you say, only works by referencing stored map data, and is only as accurate, complete and up-to-date as the provider (most likely Google) has made available. But that's mostly irrelevant to us, since terrain following only works in ArduPlane for now. They're aiming to support it in ArduCopter 3.4, but no guaranteed, and 3.3 just came out this week or last.

So, that leaves sonar or lidar, with sonar only working in the 7-15 meter range. Seems like sonar would be fine for landing, but I'm not sure how useful it would be for general flight since most trees easily reach the upper end of sonar's range. On the positive side, there are readily available sonar and lidar units that already work with ArduCopter, and mostly just need to be plugged into an I2C port and configured in Mission Planner. Not sure if the Solo has an accessible actual I2C port, or if it would need to go through the accessory port. But given that the accessory port exists specifically for this type of thing, it should be pretty easy to implement.

On the subject of my original post - landing at an alternate elevation, I've been testing it with my 250, and have found a solution in just lowering the normal descent speed to 1/2 way between the standard 150cm per second and the final landing speed of 50cm/s - I changed it to 100cm/s and it seems to land just fine, the descent from the last waypoint just takes a little longer than it used to. This is on my 250 though, so I might still go lower on the Solo if landing at 100cm/s seems too hard.

 

[RichWest]

Curious, what is the common weight for existing lidar systems designed for MR?