Flying indoors?

[Mark11]

I’d like to hear other pilot’ experiences flying indoors. I have very limited experience just 3 locations.
Location 1 - A ’black box‘ theater space 45’ x 75 ‘ with 20 foot ceilings, black walls, ceiling, floors. Obviously flying in Atti mode and the Evo was drifting, hard to correct for drift and when I tried to correct, minor control inputs would not have an effect and then suddenly produce a big over correct.
Location 2- Atrium same building as black box theater but a lit space with large windows . Drone flew well in this space, normal hover, no drift , normal response of controls, totally flyable for photographic purposes. The atrium is adjacent to the black box, just other side of the wall in the same building.
Location 3- well lit hospital atrium 3 stories tall about 70 ‘ in diameter, well lit. Drone flew much the same as the black box theater, drifting, erratic control inputs when trying to correct for drift and very borderline useable in this situation.
I thought that the problem was lack of visual cues in black box theater causing the instability when compared to the adjacent atrium Where the aircraft flew normally. Locations 1+2 were test flights prior to location 3 which was for a job. and I thought location 3 would be much the same as location 2 both being well lit.
I was wrong about that and what I am want to know now is what kind of ghosts I was fighting in locations 1 and 3 ? Its difficult to to figure out and since we can‘t see or hear radio interference, I always wonder if there is some kind of wifi or other radio situation that causes these issues in different places.

 

[herein2021]

I’d like to hear other pilot’ experiences flying indoors. I have very limited experience just 3 locations.
Location 1 - A ’black box‘ theater space 45’ x 75 ‘ with 20 foot ceilings, black walls, ceiling, floors. Obviously flying in Atti mode and the Evo was drifting, hard to correct for drift and when I tried to correct, minor control inputs would not have an effect and then suddenly produce a big over correct.
Location 2- Atrium same building as black box theater but a lit space with large windows . Drone flew well in this space, normal hover, no drift , normal response of controls, totally flyable for photographic purposes. The atrium is adjacent to the black box, just other side of the wall in the same building.
Location 3- well lit hospital atrium 3 stories tall about 70 ‘ in diameter, well lit. Drone flew much the same as the black box theater, drifting, erratic control inputs when trying to correct for drift and very borderline useable in this situation.
I thought that the problem was lack of visual cues in black box theater causing the instability when compared to the adjacent atrium Where the aircraft flew normally. Locations 1+2 were test flights prior to location 3 which was for a job. and I thought location 3 would be much the same as location 2 both being well lit.
I was wrong about that and what I am want to know now is what kind of ghosts I was fighting in locations 1 and 3 ? Its difficult to to figure out and since we can‘t see or hear radio interference, I always wonder if there is some kind of wifi or other radio situation that causes these issues in different places.

Without GPS lock these drones are always going to be hit or miss indoors. That's why I use smaller drones than the EVO indoors with prop guards. No way to predict what you are going to run into indoors. It is not Radio or WiFi interference though, those have nothing to do with drone stability, with interference you lose the signal to the drone and typically video drops out first.

The drone is probably reducing responsiveness because the EVOs err on the side of caution especially without GPS lock....they get quite sluggish as the drone tries to get a GPS signal and has to rely on its OA sensors for everything. Personally I avoid flying indoors in nearly every situation and won't hesitate to turn down a job that I consider not worth the risk to my equipment. IMO drones really are best outdoors and hundreds of feet in the air. Around here the area is saturated with drone pilots and I am quite happy to let them take the riskier jobs. Over 10yrs later many of them pop up and disappear all the time while I still have a solid customer base.

 

[Mark11]

Point taken. I had a Mavic 2 Pro and flew that drone in the black box theater in ATTI mode, and there was a switch for that on the controller. The flight characteristics of the Mavic were different but not much better in the black box theater. I wish Autel had a way to expressly switch into ATTI but apparently there is no route for that. I did fly with prop guards indoors and did not hesitate to put the drone down when it was behaving badly.
I did see some videos of EVO 2's flying indoors and they weren't bad basically normal compared to what I was experiencing.

 

[Mark11]

Youtube research yielded a video that showed a Evo 2 with tinfoil over the GPS section of the aircraft being used briefly in active track mode and while brief, the aircraft was stable. I tried the tinfoil hat on my aircraft in my yard and software indicated that navigation was in star point mode, which I interpret to be ATTI mode . Am I wrong about that?

 

[gschulzuio]

Have flown EVO IIs indoors many times at different locations of various size, height, width, lighting and other conditions, and with prop guards. Having also flown DJI Mavics and Mini's indoors, found the EVO IIs to be a bit more stable and predictable, and knock knock knock on wood, less prone to compass issues as well (ymmv). Otoh, for indoor flying have switched to a DJI Avata given its smaller footprint and ability to get into tighter spaces

Back to EVO IIs, when flying those indoors, I assume there wont be any useful GPS so will essentially be being hand flying, and to stay ahead of aircraft in terms of anticipating its behavior, drift, etc. When flying indoors I keep the aircraft fairly close and in front of me holding controller up and out in front of me so I can keep an eye on aircraft with quick glances at screen as needed. Key have found is to really keep an eye on aircraft, watch and anticipate its drift, and get it on the ground ASAP if something is not right. OA can help and hinder when indoors, on one hand when an obstacle is sensed the aircraft might pause which can be good, or on the other hand, a pain if in the middle of shooting a video sequence, or, trying to land the aircraft quickly. Thus be ready to toggle OA on/off as needed.

Something else to watch out for indoors besides obstacles are HVAC/air vents, as well as any equipment that might cause RF interference. I also have the bottom light set to flash, plus have Firehouse external lights on to help with visual cues while watching the aircraft indoors. If you are not used to stick flying, find somewhere where you can practice, even with a simulator to get used to quick adjustments yet not over correcting.

Also, get a feel for how your particular aircraft behaves so as to stay ahead of it (not physically) in terms of anticipating what its going to to, vs. having to react.

 

[Mark11]

I used to practice stick flying with my Mavic a bit and also with an inspire 1 so I am familiar with the feel. making the Evo wear a tinfoil hat sends it into star point mode which is stick flying. I had confusing experiences flying indoors, one time stable and very controllable and 2X very what I call 'slippery' lots of hard to correct drift and needing large control inputs to counteract even a slight drift which can then both counter act and then cause a large response seconds later. I don't anticipate much flying indoors but I want to be prepared if it ever comes up again.

 

[herein2021]

I used to practice stick flying with my Mavic a bit and also with an inspire 1 so I am familiar with the feel. making the Evo wear a tinfoil hat sends it into star point mode which is stick flying. I had confusing experiences flying indoors, one time stable and very controllable and 2X very what I call 'slippery' lots of hard to correct drift and needing large control inputs to counteract even a slight drift which can then both counter act and then cause a large response seconds later. I don't anticipate much flying indoors but I want to be prepared if it ever comes up again.

It is also entirely possible that your EVO has a software bug. I still have the EVO II V1 and haven't had many of the problems that I've seen posted on this forum, it seems to me that subsequent versions of the EVO got worse in the bug department.

I was flying various "things" way before drones came out and there was no GPS back then so I am very familiar with flying in manual mode; but those aircraft always had predictable responses and linear feedback; flying an EVO that goes from not responding to overcorrection would be a very disconcerting flight experience. I already encounter it at times when trying to fight with OA in tight spaces,

 

[Mark11]

Disconcerting is exactly my reaction to the no correction then over correction from control inputs. To be fair my Mavic 2 pro when flown in the same black box theater setting I mentioned did exhibit similar no correction then over correction to control inputs. I did disable EVO OA on the interior tests & flights because OA was too limiting in terms of movements that I needed to make. The tin foil hat test in my yard today saw a very stable aircraft, no surprises applying control inputs and aside from some minor altitude shifts the aircraft was very stable

 

[herein2021]

Disconcerting is exactly my reaction to the no correction then over correction from control inputs. To be fair my Mavic 2 pro when flown in the same black box theater setting I mentioned did exhibit similar no correction then over correction to control inputs. I did disable EVO OA on the interior tests & flights because OA was too limiting in terms of movements that I needed to make. The tin foil hat test in my yard today saw a very stable aircraft, no surprises applying control inputs and aside from some minor altitude shifts the aircraft was very stable

IMO these types of drones just can't get out of their own way when you need them to. They have too much automation and technology designed to make them flyable by the lowest common denominator to act in a predictable way when you attempt to disable that technology. The most predictable indoors drone would probably be an FPV drone which is designed from the ground up to be flown with no OA or GPS input. If you check YT most if not all of the best indoor drone flying videos are created by FPV drones.

I have considered getting or building an FPV drone just to relive my manual flying days, but these days I don't really have time for hobbies and I don't fly indoors professionally often enough to justify it as a business expense. If flying indoors on a regular basis provided an acceptable ROI in my area then I would almost certainly use an FPV drone to do it.

 

[uas_gian]

Have flown EVO IIs indoors many times at different locations of various size, height, width, lighting and other conditions, and with prop guards.

Yeah flying V1 Evo 2 indoors with indoor mode switched on in the app is no guarantee it sometimes misbehaves. I only crashed my drone once and it was indeed indoors. That one time it crashed into a wall because while hovering it began moving sideways for no reason and the drone ignored my counter action on the controller. I must admit I didn’t have the prop guards on that day.

So yes indoor mode gives pilots a false feeling of security for not having a gps positioning.

My post accident analysis is, the concrete structure at some position of the drone had a lot more of iron or even electrical circuit inside than I could know and this fooled the drone for its actual position. The indoor mode switch in the software doesn’t cancel this misassumption, so in my case it started moveing sideways while ignoring human input, subsequently means hit the wall and fell to the ground.

Drone and gimbal did survive the impact of about 3 feet to the concrete floor. Just one propeller needed replacement.

 

[herein2021]

Yeah flinging V1 Evo 2 indoors with indoor mode switched on in the app is no guarantee it sometimes misbehaves. I only crashed my drone once and it was indeed indoors. That one time it crashed into a wall because it kept moving sideways for no reason and ignored my counter action on the controller, I must admit I didn’t have the prop guards on that day.

So yes indoor mode gives pilots a false feeling of security for not having gps support.

My post accident analysis is, the concrete structure at some position of the drone had a lot more of iron or even electrical circuit inside than I could know and this fooled the drone for its actual position. The indoor mode in software doesn’t cancel this misassumption, so it started moveing sideways and ignored human input, subsequently means hit the wall and fall to the ground.

drone and gimbal did survive the impact of about 3 feet to the concrete floor. Just one propeller needed replacement.

People really underestimate the influence of rebar in drones especially when there are no warnings on the screen. I've tried to take off from parking garages in the past and had to place the EVO on top of its case on top of the roof of my car just to get far enough away from the rebar to not influence it. I can only imagine how bad it would be flying indoors surrounded by cement.

 

[Mark11]

The Rebar in structures is another one of those hidden pitfalls of drone operation. My Mavic 2 Pro was at times very difficult to compass calibrate. I fly frequently in an urban area with lots of tall buildings and like to fly off the tops of buildings, roof decks etc for the unobstructed view and to be away from people. I have spent half an hour trying to get the mavic to compass calibrate properly in some situations. The EVO on the other hand seems to rarely throw a compass warning though it has happened a few times when the aircraft was booted up while placed on a concrete bridge surface that was reinforced. Its easy to assume what concrete structures are reinforced - virtually all of them- but the other hidden factor is the EMF and other signals operating environment which we are bathed in and cannot visualize or understand in any direct way.

 

[MNT DRONE SOLUTIONS]

Thanks for the information fellas! I've only had a couple of indoor jobs, but never flown indoors in a no light environment. Great information for others ! Thanks for sharing.

 

[pedals2paddles]

Yeah flying V1 Evo 2 indoors with indoor mode switched on in the app is no guarantee it sometimes misbehaves. I only crashed my drone once and it was indeed indoors. That one time it crashed into a wall because while hovering it began moving sideways for no reason and the drone ignored my counter action on the controller. I must admit I didn’t have the prop guards on that day.

So yes indoor mode gives pilots a false feeling of security for not having a gps positioning.

My post accident analysis is, the concrete structure at some position of the drone had a lot more of iron or even electrical circuit inside than I could know and this fooled the drone for its actual position. The indoor mode switch in the software doesn’t cancel this misassumption, so in my case it started moveing sideways while ignoring human input, subsequently means hit the wall and fell to the ground.

Drone and gimbal did survive the impact of about 3 feet to the concrete floor. Just one propeller needed replacement.

There is no indoor mode and there is no indoor mode button or switch. This functionality does not exist on any version of the Evo past or present. This is why your drone crashed, and this is why using an Evo indoors is always very risky and subject to uncontrollable flight. It is not designed for this. The GPS cannot be disabled. Weak or multipath GPS signal will cause uncommanded and undesired movement. Rebar and electrical wiring can mess with the compass heading, not the pitch/roll.

 

[gschulzuio]

The Rebar in structures is another one of those hidden pitfalls of drone operation. My Mavic 2 Pro was at times very difficult to compass calibrate. I fly frequently in an urban area with lots of tall buildings and like to fly off the tops of buildings, roof decks etc for the unobstructed view and to be away from people. I have spent half an hour trying to get the mavic to compass calibrate properly in some situations. The EVO on the other hand seems to rarely throw a compass warning though it has happened a few times when the aircraft was booted up while placed on a concrete bridge surface that was reinforced. Its easy to assume what concrete structures are reinforced - virtually all of them- but the other hidden factor is the EMF and other signals operating environment which we are bathed in and cannot visualize or understand in any direct way.

Concur, have found the DJI Mavics to be a lot more sensitive to rebar and other metalic items vs. the EVO IIs.

 

[uas_gian]

There is no indoor mode and there is no indoor mode button or switch. This functionality does not exist on any version of the Evo past or present. This is why your drone crashed,

There is an indoor mode switch directly under the novice mode switch on the current and past few versions of iOS app. Was introduced in an update about mid 2021.

I can confirm there is indeed no indoor mode switch in the Android app that is listed as current version on the Smart Controller.

So this is getting more interesting. Why would there be a switch if this doesn’t activate or deactivate any helpful features. Moreover why are the most recent versions of the iOS app in layout different from the Android app?

The reported crash in my case was not caused by this switch, but more due to the fact the counter action on the controller to a drifting aircraft had no effect at all, or as some other pilots in this thread have confirmed with delay it would sometimes respond with an overreaction. Between me noticing the drone was moving slowly toward a wall and the impact some 2 or 3 seconds passed, should have been enough time to counter act but the actions had no effect on the aircraft. Before that drift started the drone was perfectly hovering in place for almost one minute. I was only standing 10 feet away from the drone, and the wifi feed was good.

With propeller guards on this would not have resulted in a hard crash to the floor. I have softly bumped into objects before with the guards on and the drone would just bounce back for about 3 inches. My fault for not having the guards on that day.

Questions remaining: why is there a switch labeled “indoor mode” in some cases and what else can cause the fatal delay in response?

 

[pedals2paddles]

Notice the wording. It is adjusting parameters to optimize for indoor flight. Most likely to limit sudden and high speed maneuvers, but of course it is non-specific and totally undocumented. It is NOT disabling the GPS. And in fact says it will automatically begin using the GPS once available. Which leads right into...

The reported crash in my case was not caused by this switch, but more due to the fact the counter action on the controller to a drifting aircraft had no effect at all, or as some other pilots in this thread have confirmed with delay it would sometimes respond with an overreaction.

This is weak, inaccurate, multipathing GPS signal causing the drone to react to non existent position changes. It thinks it is drifting because the poor GPS signal is resulting in false position data. It is attempting to compensate to hold position.

 

[Mark11]

Then is it better to go with the tinfoil hat indoors, to completely deny GPS when indoors? There is a youtube video of a guy in Pa, flying inside an office building the flight controls look pretty decent. He does say at one point that he is getting GPS to some degree even through he is inside and what looks to be an interior hallway.

 

[pedals2paddles]

Then is it better to go with the tinfoil hat indoors, to completely deny GPS when indoors? There is a youtube video of a guy in Pa, flying inside an office building the flight controls look pretty decent. He does say at one point that he is getting GPS to some degree even through he is inside and what looks to be an interior hallway.

No, sticking a piece of tin foil on top is not going to adequately disable the GPS.

The Evo is not designed or intended to fly safely and reliably indoors. Other than opening it up and psychically disconnecting the GPS, there is no way to disable the GPS.

 

[Mark11]

I did do a short flight outdoors with the tinfoil hat and the left hand side status check list on the controller stated navigation was Starpoint which I took to be something other than GPS.