I've been assigned the task of writing a beyond line of sight waiver app for my state agency. I've learned enough about what the DC FAA review team expects that I think I can include some areas that would require the UAV to fly behind a ridge line. Although I'm not going to expect too much and will limit how much of this drama I plan for.  We understand that most UAVs require a direct line of sight from the controller. Has anyone in the community field-verified that if, flying an "auto" mission, such as Map Pilot or Litchi,  a phantom 4 pro will continue to follow the programmed mission, even if it loses the signal from the controller?  

Here's YouTube I found for Litchi, hope it helps.

I watched that signal loss vid. Thank you. Does anyone know if the mission is continued with Map Pilot or Drone Deploy. I assume so. Also, I like to read any comments on this signal booster for Phantom. Has anyone ever used one of these signal boosters with the P4 out 3 miles?
https://www.amazon.com/ITELITE-E ... mp;keywords=Itelite

If you were to lose contact with your RC, in a flight mission, it may cause the drone to engage auto RTH.



RedHotPoker

trooper1 Posted at 12-7 11:23
I watched that signal loss vid. Thank you. Does anyone know if the mission is continued with Map Pilot or Drone Deploy. I assume so. Also, I like to read any comments on this signal booster for Phantom. Has anyone ever used one of these signal boosters with the P4 out 3 miles?
https://www.amazon.com/ITELITE-EXTENDER-ITE-DBS-MR01-2W-Range-14dBi/dp/B073XVLB7M/ref=sr_1_9?s=electronics&ie=UTF8&qid=1519175761&sr=1-9&keywords=Itelite

Any mission flown with Litchi will continue if the signal is lost, there are many members on here who will testify to that. Can't say for the other apps.

Wow!  What a task!

Just to be clear - the DJI flight app won't allow the AC to fly beyond the control signal's range.  But other apps such as Litchi, will allow it.  If you are going to write a waiver regarding that kind of a mission, it will be complicated because the drone will be on a set path and it won't be possible to vary from that path once the drone is out of control range.  In fact, once the drone is out of control range, it is almost certainly completely invisible to the pilot unless he has a fluid GPS tracker coupled with SMR capabillity (cell tower connection).  But even then, he can't control the mission or interrupt the mission.  

Flying "behind a ridge line" is certainly possible and very easily done provided the signal strength is adequate and the ridge line is relatively close.  Its possibility and range can also be enhanced depending on atmospheric conditions.  

I have some experience in this area.  And I do have quite a few different ITELITE antennas.  PM me if you want.  

Mark. Great post. Very compelling points about losing the live video feed when we lose signal and the bird flying on that path no matter what gets in the way.
That point alone screws us for getting approved, since we wouldn't be able to maneuver the bird if something flies in the way. The other pilots and I had not thought of that consequence.

trooper1 Posted at 12-7 15:41
Mark. Great post. Very compelling points about losing the live video feed when we lose signal and the bird flying on that path no matter what gets in the way.
That point alone screws us for getting approved, since we wouldn't be able to maneuver the bird if something flies in the way. The other pilots and I had not thought of that consequence.

Yes but that doesn't completely kill BVLOS because BVLOS can still be done with control and FPV.  It can be done without it being an autonomous mission.  An autonomous mission is not necessary to fly BVLOS or "behind a ridge line."  

Let's define BVLOS.  BVLOS means Beyond Visual Line Of Sight which simply means the PIC (pilot in command) cannot see the drone while it's in the sky.  That might be because it's too far to see, and it might be because it's behind an object.  

BVLOS doesn't necessarily mean loss of control or FPV.  Specifically, BVLOS means you can't glance up into the sky and see the aircraft.  But that doesn't mean you can't control it or have live video on your mobile device as you fly.  There is a lot of progress regarding fed/state regulations to be made in this area, in my opinion.  I think both Fed and state regs are a quite a bit behind the tech in this regard.  I think you have an opportunity to promote that fact.  

You're right: BVLOS is not just intervening terrain. We're going for a BVLOS waiver to fly the large, flat, sparsely populated valleys in NV, such as Clayton Valley; a large salt flat that has one of the largest lithium concentrations in NA. No need to get behind intervening terrain to fly miles beyond los.  However, even tho our agency has the need to fly a lot of remote valleys, we still have need to fly in mountainous areas. With a loss of signal control, we couldn't switch out of auto mode and manually fly it, even if our VO called in an emergency downrange.
Flying BVLOS is going to require tech to bounce the control, like the USAF does it, exc without the satellites. Maybe VOs equipped with portable repeaters. I'll have to limit our mountain missions to one's where the pilot hikes ontop of a hill and has unbroken, but distance los to the desired target area.  I'd like your opinion: do you think the FAA review team would approve us flying in a remote valley that is within a Military Operations Area, MoA?   Southern Nv has huge MoAs; these are in addition to the Restricted airspaces, that we won't propose to fly in. At this point, I'm considering dropping the half dozen BVLOS candidate areas within MoAs.

trooper1 give Dirty Bird a ping (https://forum.dji.com/thread-119703-6-1.html) . He, as have others, has been using Litchi on long flights for some time.

My personal opinion is I don't think FAA will approve any flight where the PIC can't see the aircraft.  I'm not aware of anything like this ever being approved by FAA.  There were media articles last fall 2017 about this kind of thing, trial runs with FAA oversight being made out of VLOS sort of as an experiment.  I think they did 2 miles (haha).  But I haven't heard of much happening since then.  

Here's just one article.

https://www.washingtonpost.com/l ... _term=.6585ba881a9d

Trying to get it approved in an MOA seems about impossible to me.  As you know, you've got the potential of those fast movers at low altitude.  You'd have to know the schedule of those jets.  

I do like the mountain top idea, but not in an MOA.  One could fly amazing distances with full control and non-lagging real-time high-def FPV from a mountain top.  Ten miles out and back can be done.  Twelve too.  Fourteen miles can be done.  Sixteen miles out and back can and has been done - even without the mountain top - operating a modified P3 or P4 consumer controller/drone and flying out from flat terrain.  It's documented.  

Also, the repeater (relay thing) could work.  There was a guy who made those for P2s a few years ago and they actually worked well.  He put the repeater on a hovering drone which would stay stationary overhead and then sent the second main P2 out for a distance - well "beyond the ridgeline" of the PIC and everything worked well.  

Good luck

Mark, thank you for your opinion about proposing MoAs in my waiver app. I'm going to do as you advise and drop the candidate areas within MoAs. We're going to take an operational hit not be able to fly BVLOs in  those areas, but I can always write another waiver just for those areas later on.  I read your Post article. Actually, 23 waivers (107.31) have been approved by the FAA [out of only 2,000 waiver requests for BVLOS). I've read every one of the 107.31  waiver approval certs at this page: https://www.faa.gov/uas/request_waiver/waivers_granted/  I've been advised that, due to the nature of Nevada topography and that a gov agency is requesting, FAA might be more open to approving a BVLOS waiver. Of course, I've learned the FAA is not happy until remote pilots are unhappy. FAA guys have told me that we'd be the first gov agency. Some of you might not be aware that Precision Hawk is so confident that they can get a BVLOS waiver approved for a client, that they're charging $30K to prep the wavier. Rather than paying the $30K, my state mangers have decided better that I should expend the 120 hours to figure out and write a waiver. I think most tax payers would agree paying me for 80 hours of work is a huge saving over cutting a gov check to Precision Hawk for $30K. However, given the frustration I've experienced dealing with the FAA to this point, some of you might opt to cut the check...
  Mark, can you think of anything that might help me track down the guy that did the repeater work? Or the signal repeater? I'll research it and post back what I find. That's a compelling idea that I need to follow up on

That's interesting about the BVLOS waivers being granted!  I didn't know that.  One percent is better than nothing!

The guy who did the P2 relays is here:  http://redbirdrf.com/

But his work is with the more fundamental control signal and wifi signal of the P2 and also the P3S.  Lightbridge and occusync are different systems so it might not be possible to do what he did with the P3 Lightbridge and above.  

Good luck!

I've researched the Redbird repeater; talked to the guy than markets it. The problem is that it only works for the Phantom 3 Standard.   Here's an interesting vid about bouncing the control signal:  start the clip at 12:57     Does anyone know of a repeater that works for Phantom 4 Pro?  

I didn't think the state could make such waivers? That is totally the FAA jurisdiction.  

DJI seems to be 3 steps ahead of the rest of the world when it comes to innovation, and I wouldn't be surprised if DJI is already working on such a relay.  They may have already finished it and are waiting for the world's laws to catch up.  But the modern DJI systems are proprietary (i.e. secret) and so it's going to be difficult to develop something like the redbird system and apply it to the modern DJI systems.  

There are also other drone companies who may be making drones with tech which is more compatible to a relay.

There is also the SMR (cell tower connection) possibility someday.  

There are also ways to do it without a relay.  For example, in a city, deliberately bouncing the signals off a nearby building could allow control and FPV behind a second building out of VLOS.  Even the "below ridgeline" thing could work if the ridgeline and drone were close enough and if the transmitters were strong enough and the angle beyond the ridgeline was shallow enough.

Thank you for confirming that there is no current repeater or relay available for the newer DJI products. I'll follow your advice to research other UAV models which might offer a relay option; I'd not thought of that. To clear up a point: I'm preparing a waiver application, on behalf of the my agency, for the FAA

Does anyone know of a failure study for the Phantom transmitter. A statistical study that predicts the failure rate of the transmitter during flights? The FAA expects to see such data in the waiver apps.

trooper1 Posted at 12-7 15:41
Mark. Great post. Very compelling points about losing the live video feed when we lose signal and the bird flying on that path no matter what gets in the way.
That point alone screws us for getting approved, since we wouldn't be able to maneuver the bird if something flies in the way. The other pilots and I had not thought of that consequence.

If the drone is a P4 or higher the drone has obstacle avoidance, and I assume even without RC connected it would behave in a similar fashion to an incident I almost had when I first got Litchi. I made a mission for my drone to fly around the local high school baseball and football fields, soccer too, tennis courts...There was a total of about 6 fields so lots of good POIs for me to plan my mission....When I had made the mission, the satellite image showed the field i was taking off and landing from was clear of any obstacles....Well as you can probably guess it wasn't.  I know I should have checked to make sure nothing was on the field, but hindsight is 20/20. There just happened be a soccer goal posts and net in the middle of the field. Didn't notice it taking off, but my last waypoint of the mission was maybe 20ft from the take off point and in direct path with soccer goal. I had it come down in various altitudes using the last few waypoints. Of course the last waypoint was at a height of 5 ft off the ground. So as its coming toward me I see the goal getting closer and closer and I almost sh*t a brick. Thank god for the obstacle avoidance, because it stopped right at the post, then it went around the post. I was so relieved my drone made it safe. My lesson learned (luckily without any damage): ALWAYS check the areas the drone will be flying and even triple check if you have to and make sure there are no obstacles of any kind.

I've never heard of such a study, but if I was to guess based on experience and reading the forums, I'd say the RC failure rate during flight would be something like less than one one-hundredth of 1% excluding pilot error (e.g. low RC battery, RC dropped on concrete, dog leaps up and bite antenna off, etc).  

Mark,
  I appreciate your opinion about the absence of studies. I think so too, but I've got to keep researching. The FAA is not going to accept our failure estimate. 2,000 BVLOS submitted - 23 approved. The faa says the main reason for rejecting a bvlos app is "lack of information" applicants making "assumptions".   The FAA is not happy until pilots are unhappy. No wonder Precision Hawk can  charge clients $30K to prep a bvlos app.
How many people on here want to pay $30K for a waiver?
I'm hoping someone from DJI will advise me on failure studies.

This is one really awesome thread and everyone's contributions as replies to your starting point are outstanding.

I would appreciate anyone replying back with their experience of an actual transmitter failure and the effect on their P4. If the on board gps failed, would the barometric pressure gauge allow the P4 to hover in place?

trooper1 Posted at 12-24 08:53
If the on board gps failed, would the barometric pressure gauge allow the P4 to hover in place?

That's an easy one.  If the GPS fails, the AC would fly in "ATTI" mode.  The altimeter would not be affected.  The AC could be safely flown home.  But if the GPS fails and the RC control connection was broken at the same time. the AC would hover as long as the battery lasted.  But it would drift with the wind.  Eventually it would auto-land if it didn't hit something first, such as a tree or building.  Also, on the P4s, I believe they have redundancy systems.  Somebody else might want to comment on that.  

Mark, thank you for the GPS failure post. After some hours of research, with no better results, I may have to just submit, to the FAA, our reasonable estimate of the transmitter failure rate. I really wish a DJI employee would address my question:  are there any formal failure analysis for P4. Has anyone confirmed that the lights from two lume cubes can be seen 3 miles away, day or night?  https://lumecube.com/  2 lume cube lights, mounted on a Phantom would put out 3,000 lumens. I can't find any documentation to verify that 3,000 lumens would be visible at 3 miles.