What if DJI set a RTH option which allows the drone to "back-track" home by following basic waypoints recorded along the flight path.
It would not replace the current RTH options, but would merely be another option added to the list.., to be used at the pilot discretion.

For example. I generally leave my 'RC Signal Lost' setting on RTH, at 120m as most interference comes from trees, or small hills. If signal is lost, the drone rises above the blockage to regain signal well before it reaches 120m.., and I can fly back to a safer location. I prefer to keep the maximum 120m RTH height, because sometimes an obstacle may be higher than anticipated, so it's better to rise above, than crash.


The only thing I must keep in mind is the potential of a complete RC failure, with a low battery. An unstoppable 120m climb could cause battery problems.
Occasionally I will set my 'RC Signal Lost' to HOVER when pod-racing around the bush under tree canopies. I can usually walk to regain signal.
I never use the LAND setting.

Having a basic waypoint RTH option would be beneficial for most open/outdoor flying environments.

This would help reduce crashes, where people fly around a building or hill.., lose signal.., and RTH height is too low. Instead of flying it a direct line home.., drone would merely stop, turn back, and follow the safe path it took to reach its current location.., either till it reached HP or signal is regained.
It could also help reduce drones from flying upwards into trees and ceilings as a reaction when experiencing various problems.
As always, it would be up to the pilot to choose when to use the correct setting.

I like the mission is completed his route after signal lost. This is a planned, orgenized and safe route (because i flied or walked it) so no RTH interference.
(only on low battery)

But this would open doors to different kind of scenarios, and in case of RC signal lost how you could set waypoints?
Maybe it could be good idea if they allow the RTH by flying the inverse track that you already done (if is not a dynamic environment)

I am using FPV camera for waypoints. And i am always checking the route before i fly. But when i start the mission i am sure that any interference of the RC signal not stop the mission - so i always know where the drone is. But i never lost my signal in my missions so far - but for me it is one unreliable factor less.

For manual flying i use always DJI app and happy with the RTH funtction as it is.

Planning a RTH back route is possibe but no need for me.

The Litchi app does this already.  You can fly the waypoint  route and make the speed a negative number at the end to fly the reverse.

Nicola90s Posted at 2016-2-15 07:07
But this would open doors to different kind of scenarios, and in case of RC signal lost how you coul ...

New scenarios.., yes. This is not a bad thing IMO.
The waypoints would be created by the drone.., during the flight. Not the pilot.
Yes. There may be a new obstacle in the flight path... But if you think about what objects would suddenly appear along your flight path (another drone)., the risk is no different to any other automatic RTH function where it would simply fly up where there may also be another drone.
A risk analysis would be good to show potential hazards with the idea.
Then compare that with potential hazards and known crashes caused by the current RTH protocol to see which is safer.., if there can only be one that is.
I think it's not one or the other.., but perhaps a combination of options.

SHamers Posted at 2016-2-15 07:29
I am using FPV camera for waypoints. And i am always checking the route before i fly. But when i sta ...

It would not replace the current RTH options.
Although the current RTH procedure is sufficient.., and I am also happy.., i think this is a great idea to help improve safety.
Especially after reading about so many RTH crashes.., it seems like this would solve most of those issues.It's basically the current RTH feature.., but instead of going straight up..., it would try flying in reverse to regain signal.

So,  to be clear,  you are talking about doing a flight and periodically setting waypoints along the way? And then pressing rth button and having the drone follow those points on the return path?  As it exists,  that would not work.  Because it would go to rth height and make a beeline straight back.

Kneepuck Posted at 2016-2-15 09:15
The Litchi app does this already.  You can fly the waypoint  route and make the speed a negative num ...

Oh, ok. Perhaps DJI would also like to adapt an automated RTH waypoint protocol.
Does Litchi allow you to choose the automated waypoint RTH option?

Can anyone upload an example of Litchi performing this RTH procedure? I'm interested in how/where it sets a "safe" point, and how many are set, and how often.

Kneepuck Posted at 2016-2-15 09:15
The Litchi app does this already.  You can fly the waypoint  route and make the speed a negative num ...

Oh, ok. Perhaps DJI would also like to adapt an automated RTH waypoint protocol.
Does Litchi allow you to choose the automated waypoint RTH option?

Can anyone upload an example of Litchi performing this RTH procedure? I'm interested in how/where it sets a "safe" point, and how many are set, and how often.

Ok,  I see I misunderstood your initial post,  mostly the automated waypoint thing.  What you are talking about is for Dji to add a new feature to the existing Dji-Go app.  You got me there,  I do know that Dji does add features from time to time,  but it looks like these are features are usually ones some other drone comes out with and Dji then puts in the Go app.  They may accept suggestions from actual users and use them,  but I have yet to see that.  And frankly,  the Dji app has unresolved issues already that need to be addressed,  but never seem to be.  It would be great if there were a suggestion site that was actually used by Dji,  but for the time being,  the only way to get this type of thing done is write your own app.  

Kneepuck Posted at 2016-2-15 09:52
Ok,  I see I misunderstood your initial post,  mostly the automated waypoint thing.  What you are ta ...

Thanks. Maybe the idea could be passed on to the DJI SDK community.

Kit Walker Posted at 2016-2-15 10:01
Thanks. Maybe the idea could be passed on to the DJI SDK community.

I understand what you are asking, and this topic was discussed at length on another thread quite recently. One of the issues is that when RTH is invoked for low battery, the aircraft needs to fly back by the most direct route, otherwise it risks running out of battery before it gets home.

Yes, DJI could offer it as an option, but as it puts the aircraft at greater risk of loss, you might understand why they are not interested in the idea.

Geebax Posted at 2016-2-15 16:53
I understand what you are asking, and this topic was discussed at length on another thread quite re ...

Thanks. I haven't been on the forums for a long time.
I think the the shortest route home is great for a low battery.., but not for signal loss. I believe it would be quicker to backtrack in certain/most situations.


Although, looking at it from a critical battery perspective.., it may work out more energy efficient to travel horizontally, rather than vertical in some circumstances.
E.g. My RTH height is always set to 120m standard.., especially when entering new terrain.
If my bird was 2km away, and decided to RTH.., it would take X amount of minutes to ascend the 120m.
That time could be spent flying horizontally (if the pilot deems safe.., just like any RTH feature).

So in certain situations.., I believe this would be better.
Signal loss situations..., where the pilot does not want to ascend during signal loss (tree, powerlines, ceiling, wind, predator etc)., nor land (water, dust, uneven).., nor hover (not enough time to walk and regain signal.., thus bird ascends into hazard).

Kit Walker Posted at 2016-2-16 04:37
Thanks. I haven't been on the forums for a long time.
I think the the shortest route home is great  ...

If I remember correctly, this arose because a forum member flew around behind a huge building in Miami, lost the signal, the aircraft went into RTH and tried to fly home through the building. This is a scenario that happens from time to time. In which case, it would have been safer to backtrack along the original route.

Where this comes unstuck though, is when the pilot took a long and leisurely route to get behind the building, and to follow that route precisely might have cost dearly in battery time. If the original flight route was very complex, it might use a lot of battery power to re-create that route.

My opinion? DJI are damned if the do and damned if the don't. The whole RTH thing is very complex issue and it is not hard to poke holes in the methodology.

Geebax Posted at 2016-2-16 08:58
If I remember correctly, this arose because a forum member flew around behind a huge building in M ...

If the pilot preferred to use the Waypoint RTH feature to automatically bring the drone all the way back to Home Point.., the software would make allowances for the distance travelled, compared to battery life.., just as it does now.

When I ride my dirt bike (which has no fuel gauge).., I set the fuel to ON. It will cut out around half of an empty tank. I know that as long as I turn back when I hit reserve.., and take it easy with the throttle.., and do less exploring.., I will have enough fuel to return.

DJI GO already has a battery life/time/distance algorithm in place, which could be modified to suit this RTH feature.
Although minor problems like this could easily be overcome (not a reason to dismiss the idea) it would be most handy for every other situation where signal loss is the problem.., not battery life compared to distance from Home Point. The current RTH features are riddled with problems.., yet the lack of perfection does not mean the ideas should not be applied.., as long as consumers are educated on the procedure.

Kit Walker Posted at 2016-2-16 09:28
If the pilot preferred to use the Waypoint RTH feature to automatically bring the drone all the way ...

In principle I agree. But in the case I mentioned, he was not using waypoints. Just flying around the harbour.

The problem with RTH is that everyone has an idea to improve it, and DJI are not going to implement them all. For my part, I don't care, as I treat it as a flying camera. It never gets out of my sight, and I always fly back with plenty of reserve battery to land safely.

Your analogy of the dirt bike is a good one, but how does the Phantom know how to 'do less exploring' on the way home to reduce fuel consumption? It can only retrace the original flight path, and if the pilot was noodling around on the way to the point that RTH was initiated, then it must re-create all that on the way home.

Geebax Posted at 2016-2-16 09:57
In principle I agree. But in the case I mentioned, he was not using waypoints. Just flying around  ...

Using that recent crash as an example (I think it was like this sorta).., the drone would have just travelled back no problem because it was only 1km away, or under. Plenty of time to get back, even in strong wind.

If the drone did lots of zig zags along the way, instead of a straight line.., you could use GPS to calculate distance to give an average battery consumption for the flight time.., perhaps even taking into consideration the cell voltages.., to create an algorithm which would tell the drone;
IF> In WRTH (waypoint RTH), use power consumption algorithm X

Even if the drone loses signal behind that building/hill/tree/interference.., with say 40% battery at 1km away.., then it would it would be just as quick for the drone to just STOP.., and turn back.
After about 5 seconds.., the drone would regain signal.

This is much better than the RTH kicking in, and hoping you knew the exact height each time.., and potentially being wrong, and crashing.
A strategic retreat would overcome the risks immediately.
Perhaps DJI would also like to allow pilots to change the 'RC Signal Lost' options DURING flight (providing RC and bird are now connected).The pilot could then initiate the traditional RTH in a straight line procedure.


So, as the pilot flies from TOP of the beach, along the coast.., the drone would be making waypoints every 3-10 seconds (or whatever).
When signal is lost.., it would stop (as normal), and initiate a return back to the previous automatic waypoint that was set.

Kit Walker Posted at 2016-2-16 16:21
Using that recent crash as an example (I think it was like this sorta).., the drone would have just ...

Hmm, yes. A 'Strategic Retreat' movement might be a good way to tackle it, just back up and see if you regain signal, rather than try and fly home. Not quite sure how you would do that though, not an easy scenario for the aircraft to recognise.

Geebax Posted at 2016-2-16 16:40
Hmm, yes. A 'Strategic Retreat' movement might be a good way to tackle it, just back up and see if ...

Perhaps just have 2 settings.
One called 'Failsafe Mode' (where tradional RTH procedure can be selected. Activated by RTH button, and critical battery warning)
One for 'RC Signal Loss' (Where automated waypoints RTH could be initiated. Activated by signal loss)

There are waaaaaaay too many threads regarding the current RTH system.

Truly,  the only practical implementation for any rth feature is one that uses autonomous obstacle avoidance.  And some of the next generation drones to be releases claim to have just that.  It would require more sensor hardware,  so probably our P3's will not get the feature via firmware update.  I guess we will have to wait and see what happens.

Kneepuck Posted at 2016-2-18 09:16
Truly,  the only practical implementation for any rth feature is one that uses autonomous obstacle a ...

Collision avoidance will be good for RTH. Not sure if it's the only idea though...

Improving the RTH is still on the books IMO, regardless of collision avoidance technology or not. It seems hard to discuss the idea on this site for some reason.

I know I'd prefer to leave my 'RC Signal Lost' set to a Waypoint Return Option (where the drone would backtrack until signal was regained). Which.., for most worst case scenarios where crashes occur.., would be quicker to regain reception as opposed to climbing 80m+ over a skyscraper.

Just an observation.  I had a scare today with some high winds and the Litchi app crashing.  The only way I could get the app to come back was to momentarily turn off the remote control,  then reload the app after turning it on.  This,  naturally,  reset the home point to about 1000 feet away from where I was.  I had to bring it back against a strong wind,  and had 17% battery when I landed.  No amount of obstacle avoidance would have helped in that case.  I was lucky to get the P3 back in one piece.