Hello.  Since this is my first drone, I searched best drones and all the reviews point to the DJI Phantom 4 PRO.  I was going to get one, and then I heard from a friend that the 4 cannot be tinkered with, only the 3 can take other cameras and equipment.  Is this true?  If so, I will just buy the 3, so that I can put aftermarket cameras such as multi-spectral, or LiDar unit.  Thank you for your help, I  apologize for the broad questions.

AK

You might have a look at these guys in NL.  They seem to have mostly FLIR but might be able to help you out.
http://dronexpert.nl/product/pha ... n-on-own-phantom-4/

You need to didh out BIG BUCKS for this setup upward of $10,000

Do you know of a Drone/LiDar recipe that is already proven?  I have put some research into after market LiDar units for drones and they are not cheap, you are correct

If you can afford a LiDar set up, you should be looking at industrial platforms, like the M600

shawn_ Posted at 2017-2-22 11:52
If you can afford a LiDar set up, you should be looking at industrial platforms, like the M600

Exactly, modifying a Phantom is the wrong way to tackle it, they are not designed to be modified like that.

shawn_ Posted at 2017-2-22 11:52
If you can afford a LiDar set up, you should be looking at industrial platforms, like the M600

this makes complete sense.  just the way i needed ti be steered.

Geebax Posted at 2017-2-22 14:16
Exactly, modifying a Phantom is the wrong way to tackle it, they are not designed to be modified like that.

thank you!

akarrick99 Posted at 2017-2-22 20:50
thank you!

Another point is that the lightest LIDAR unit is the YellowScan® LIDAR Mapper system, and it weighs in at 2.1 Kilos, which is way outside the lifting capability of the DJI consumer drones. Just one of the integrated drone packages carrying this unit was priced at $115K

I tried fitting a small Sony 20Mp camera to my P3P.
When flown it was constantly giving "Motor Overload" warnings, and was very unstable in flight.
Camera weighed only 150g and my mounting a further 35g.
I would not reccomend mounting anything over say 100g to P3.

liningiv Posted at 2017-2-23 06:02
I tried fitting a small Sony 20Mp camera to my P3P.
When flown it was constantly giving "Motor Overload" warnings, and was very unstable in flight.
Camera weighed only 150g and my mounting a further 35g.

and you changed the C.G. down down and down......

Take a look at this camera: http://scanse.io

Bullflyer Posted at 2017-2-23 08:44
and you changed the C.G. down down and down......

So would you have raised the CofG ?

Bullflyer Posted at 2017-2-23 08:49
Take a look at this camera: http://scanse.io

will let you know if I end up getting this attachment.  I am actually leaning towards a DJI phantom 3 now instead of 4!

liningiv Posted at 2017-2-23 06:02
I tried fitting a small Sony 20Mp camera to my P3P.
When flown it was constantly giving "Motor Overload" warnings, and was very unstable in flight.
Camera weighed only 150g and my mounting a further 35g.

So could you do the same thing with a Gear 360 VR camera?

Geebax Posted at 2017-2-22 21:44
Another point is that the lightest LIDAR unit is the YellowScan® LIDAR Mapper system, and it weighs in at 2.1 Kilos, which is way outside the lifting capability of the DJI consumer drones. Just one of the integrated drone packages carrying this unit was priced at $115K

I have seen LiDar equipment cheaper than what you are speaking of, Google uses the best of the best LiDar from a company called Velodyne and the most expensive units are $80,000, they take >700,000 scans per second.

This is a cheap scanning LiDar that takes ~1,000 scans per second.  It is only $350, and comes on the market next month.  If I find a drone that is open framework and can carry it, my goal is to use it.

http://scanse.io/

akarrick99 Posted at 2017-2-26 13:35
I have seen LiDar equipment cheaper than what you are speaking of, Google uses the best of the best LiDar from a company called Velodyne and the most expensive units are $80,000, they take >700,000 scans per second.

This is a cheap scanning LiDar that takes ~1,000 scans per second.  It is only $350, and comes on the market next month.  If I find a drone that is open framework and can carry it, my goal is to use it.

OK, good luck.

akarrick99 Posted at 2017-2-26 13:35
I have seen LiDar equipment cheaper than what you are speaking of, Google uses the best of the best LiDar from a company called Velodyne and the most expensive units are $80,000, they take >700,000 scans per second.

This is a cheap scanning LiDar that takes ~1,000 scans per second.  It is only $350, and comes on the market next month.  If I find a drone that is open framework and can carry it, my goal is to use it.

That's a kickstarter project - not a product that's on the shelves and available now.
Perhaps they will deliver but having watched the Lily drone and Plexidrone projects from the sidelines , I wouldn't give a cent to a Kickstarter project.

akarrick99 Posted at 2017-2-26 13:31
So could you do the same thing with a Gear 360 VR camera?

I would not fit anything to the P3P.
It may have the initial lift capability, but how long the flight time is, and how unstable and therefore unsafe it may be I'll leave up to your trials.  I was not willing to risk the £1000 I spent on my P3P by experimenting any further.
Then guess what?  1 year later the P4P comes fitted with a 20Mp camera.  Job done, and safely.
M200 seems like what you need, but reading between the lines, is not what you want to pay?

akarrick99 Posted at 2017-2-26 13:35
I have seen LiDar equipment cheaper than what you are speaking of, Google uses the best of the best LiDar from a company called Velodyne and the most expensive units are $80,000, they take >700,000 scans per second.

This is a cheap scanning LiDar that takes ~1,000 scans per second.  It is only $350, and comes on the market next month.  If I find a drone that is open framework and can carry it, my goal is to use it.

With 1000 scans per second, I think this is good for collision avoidance and AI navigation but not ideal for mapping/surveying. You'll need more points for any real accuracy.

Labroides Posted at 2017-2-26 22:46
That's a kickstarter project - not a product that's on the shelves and available now.
Perhaps they will deliver but having watched the Lily drone and Plexidrone projects from the sidelines , I wouldn't give a cent to a Kickstarter project.

Ah, yes I did hear about the Lily failure, not the plexi though.  It says March, so I was going to see how the release went and play it by ear.  If you have any suggestions for low price lidar please let me know

liningiv Posted at 2017-2-27 03:54
I would not fit anything to the P3P.
It may have the initial lift capability, but how long the flight time is, and how unstable and therefore unsafe it may be I'll leave up to your trials.  I was not willing to risk the £1000 I spent on my P3P by experimenting any further.
Then guess what?  1 year later the P4P comes fitted with a 20Mp camera.  Job done, and safely.

correct.  I am still in grad school, and not working right now so I do not have the coin for anything more than the P 4.  Let me know if you would do anything else in my shoes

shawn_ Posted at 2017-2-27 09:02
With 1000 scans per second, I think this is good for collision avoidance and AI navigation but not ideal for mapping/surveying. You'll need more points for any real accuracy.

So could one develop their own AI navigation system using this type of device?  I do not know anything about playing with DJI's software, I doubt anything like the API is even developable outside their walls.  

How many points do you need to ensure a rigorous accuracy?  Google/Uber use 700,000 per second or more

akarrick99 Posted at 2017-2-27 22:24
So could one develop their own AI navigation system using this type of device?  I do not know anything about playing with DJI's software, I doubt anything like the API is even developable outside their walls.  

How many points do you need to ensure a rigorous accuracy?  Google/Uber use 700,000 per second or more

I've had mentioned this on one of your other threads. The Phantom 4 series is only capable of using Mobile SDK and not Onboard SDK. You can create an app that can support the P4, perhaps you can make a basic navigation app for the aircraft in that aspect, but you would not be able to perform serious configuration on how the sensors behave.

DJI-Jamie Posted at 2017-2-27 23:36
I've had mentioned this on one of your other threads. The Phantom 4 series is only capable of using Mobile SDK and not Onboard SDK. You can create an app that can support the P4, perhaps you can make a basic navigation app for the aircraft in that aspect, but you would not be able to perform serious configuration on how the sensors behave.

thank you very much for the additional information

akarrick99 Posted at 2017-2-27 22:24
So could one develop their own AI navigation system using this type of device?  I do not know anything about playing with DJI's software, I doubt anything like the API is even developable outside their walls.  

How many points do you need to ensure a rigorous accuracy?  Google/Uber use 700,000 per second or more

'I doubt anything like the API is even developable outside their walls.  '

There are several 3rd party app developers who make alternative software for flying DJI aircraft, so I guess they found the API fathomable.

akarrick99 Posted at 2017-2-27 22:23
correct.  I am still in grad school, and not working right now so I do not have the coin for anything more than the P 4.  Let me know if you would do anything else in my shoes

I don't fully understand what it is you are trying to achieve.  But it seems like the hardware nor the software is available at your price point.  And at the same time you are trying to fly.  What about one thing developed first, on the ground, then try to fly it when you have developed the camera/software?

Geebax Posted at 2017-2-28 00:22
'I doubt anything like the API is even developable outside their walls.  '

There are several 3rd party app developers who make alternative software for flying DJI aircraft, so I guess they found the API fathomable.

I was just speculating, and yes that is totally different situation for someone with much deeper knowledge of Unmanned aircraft API's than me.  Also, could you please direct me to these 3rd party developers?  Maybe I can learn something from them

liningiv Posted at 2017-2-28 01:00
I don't fully understand what it is you are trying to achieve.  But it seems like the hardware nor the software is available at your price point.  And at the same time you are trying to fly.  What about one thing developed first, on the ground, then try to fly it when you have developed the camera/software?

I do not necessarily know exactly what I am trying to achieve as well, I am new to this space but I am trying to find a drone that will give me an edge in several speculative areas.  Seems like I will start flying with DJI 4 pro and save my coin while I practice and learn

akarrick99 Posted at 2017-2-26 13:31
So could you do the same thing with a Gear 360 VR camera?

If you have access to a 3d printer you can definitely mount a 360 camera on a P4 platform:

http://www.thingiverse.com/thing:2019114

I've modified Bernd's original design to fit my P4P. My model of the payload bridge can found here:

http://www.thingiverse.com/thing:2093489

I'm still working on testing my final setup (with a shortened adjustable mount and a 360fly camera) and I'll post results on my project page...

The scanse may be the next payload I'll try, not sure what practical use it would serve though, you going to map caves ?

Spirytus Posted at 2017-2-28 13:58
If you have access to a 3d printer you can definitely mount a 360 camera on a P4 platform:

http://www.thingiverse.com/thing:2019114

Links aren't working correctly. Anyway, here's a link to the P4P payload bridge:

http://www.thingiverse.com/thing:2093489

You will see the similarities between this mount and mine, above, excepting the carbon fibre anti vibe mount that my solution included.
But the proof of the item is to fly it.  I should have just found the parts weight,and then mounted a similar weight to my P3P to find out if it would safely fly with that sort of weight.
It would have save me a lot of effort and a little money.

akarrick99 Posted at 2017-2-28 12:01
I was just speculating, and yes that is totally different situation for someone with much deeper knowledge of Unmanned aircraft API's than me.  Also, could you please direct me to these 3rd party developers?  Maybe I can learn something from them

Try Googling Litchi.

akarrick99 Posted at 2017-2-28 12:02
I do not necessarily know exactly what I am trying to achieve as well, I am new to this space but I am trying to find a drone that will give me an edge in several speculative areas.  Seems like I will start flying with DJI 4 pro and save my coin while I practice and learn

http://www.dji.com/newsroom/news ... nterprise-solutions

Bullflyer Posted at 2017-2-23 08:49
Take a look at this camera: http://scanse.io

I'm a Geomatics student working on my senior project and bought a P4 for creating orthophotos and point clouds.   If you just want to get a "rough" idea of something, and poor quality is acceptable, get the scanse.  If you want something that can produce an accurate point cloud check out the Velodyne Puck (http://velodynelidar.com/vlp-16.html).  It's more money ($8k), weighs less than 1kg and does 300k points/second and also has multiple returns (ie: can help to see "through" tree cover to the ground).  I would venture to say creating a point cloud using photogrammetric analysis would be superior in quality compared to the scanse (ie: using the built in camera and aerotriangulation to calculate x,y,z values would be denser than the scanse can do).  

Now if you just want to play and see, go for it much cheaper but the precision and accuracy are incredibly limited.  Ultimately I think you'll get better results (and way more dense point clouds) using the modern computer vision applications with traditional aerotriangulation like Pix4d, Photoscan and even the online stuff like DroneDeploy.  Lidar is neat, but to get high precision you need a high dollar.

And if you want some real impressive data, couple that Velodyne with a RealEarth stencil....But like JBin said, high precision is high dollar. And the problem with this developing technology is that 6-12 months from now, there will be faster, smaller and cheaper options...

liningiv Posted at 2017-2-23 06:02
I tried fitting a small Sony 20Mp camera to my P3P.
When flown it was constantly giving "Motor Overload" warnings, and was very unstable in flight.
Camera weighed only 150g and my mounting a further 35g.

Have you tried removing the main camera first,, then attaching just the Sony camera?

Hoping to attach my iPhone 12 Pro Soon to my Phantom 4 Pro, should be strong enough... kinda heavy but gonna try! Wish me luck! Will not work from very far but has LIDAR

You will see the similarities between this mount and mine, above, except the carbon fiber anti vibe mount that my solution included.
But the proof of the item is to fly it.  I should have just found the weight of the parts,and then mounted a similar weight to my P3P to find out if it would safely fly with that sort of weight.
It would have save me a lot of effort and a little money.

- TechDoa
https://techdoa.com

https://techdoa.com