Hello Everyone,
When I stumbled upon these amazing machine I got very excited. I do forensic animations/laser scanning/video analysis and this machine would be a wonderful tool to help us bring our work even further into the 21st century. Would be grateful for answers to my questions below.
1) How do you control the camera remotely to take a picture when it is in the proper position?
2) Is there a way for you to see what it sees in real time?
3) 4 wings vs 6 wings what are the advantages?
4) Generally what are the top speeds of these machines when you have a camera (goPro)?
5) Where get I purchase a fully built machine this month?
6) What will be the hardest challenge in learning to fly this machine.
Just to give you an idea of what I would love to do with this machine. imagine we need to animate an accident that happen on a busy freeway. Normally we would create a 3D model of the freeway, and the vehicles involved. With this machine we could capture video flying over the freeway, camera track the video into our 3D software, digitally remove all the cars in the video which would give us video of an empty freeway. Now we could simply animate into the video the vehicles that were involved in the accident and show the jury what happen from the sky.
Jason
Tags:

Permalink Reply by Michael Pursifull on September 13, 2011 at 12:48pm Jason,
I'm new here too, but I've built four of these "amazing machines" so far (three of them have even flown!) and being new, I hope I can answer some of your questions without you taking me too serious (and without discouraging others from answering.)
These are just my opinions, and I'm *maybe* 60 days ahead of you in this..
1) How do you control the camera remotely to take a picture when it is in the proper position?
As this is a DIY (do-it-yourself) community, the answer really is "however you would like." But I think what you are seeking is that many people use a 9 gram servo and servo arm fixed (tape, velcro, wedged, or otherwise) to the shutter. But some setup their camera to take timed continuous shots. Some use cameras with USB-accessable commands (existing or modified firmware ... thinking of a Canon, details in the forums) and some (like me) buy used cameras, tear them apart, and attach opto-isolated relays to arduino-mini-pro boards, and then attach those relays to the shutter electronics inside the camera ... or at least that is my plan, I have the cameras and controller, but I haven't gotten that far just yet...
That is the "shot" part, but as to the "pointing" part of control, in ArduCopter from about 2.0.37 or maybe a bit before, camera stabilization is enabled by default. So if you add a servo (and mount) for roll (and another for pitch) then you get stabilization of the camera (moves to compensate for air frame movements) and you can bind your transmitter to add input to roll and/or pitch, moving the camera (while retaining the relative compensation) Actually doing this requires a bit of work, selecting or building a mount, picking servos, and tuning it. And you might need some help with changing the firmware to get it to work they way you want. Or you can buy a camera system/mount/gimbal/ball mount - as I'm sure you are aware, you can spend as much as you want on these.... $300 - $80k or more [with many commercial camera systems being at least $3000, that I have seen, and you can buy a house for the cost of a laser guided, gimbal-lock-immune thermal object-tracking systems...]
2) Is there a way for you to see what it sees in real time?
Many people call this FPV (first person view) in general, but the concept applies, I think, more to general RC flying, with a view towards experiencing the flight as if you were an onboard pilot. However, I do have a prototype more like what you are talking about. I have a series of cameras that I purchased off of Amazon, used, to evaluate for still and video camera use, outside of the FPV system I have. Three of the four have video cables, and I have tested cutting up the cables and channeling the LCD display through my FPV transmitter. It works very well. I narrowed the cameras I will use first to prototype to a Sony Cybershot, because I have two, and because they support turning off the auto-off timer (two other models do not, so unless I provide input every few minutes, they turn off. Not my idea of fun if it is 400 feet in the air...)
3) 4 wings vs 6 wings what are the advantages?
I assume you are talking about ArduCopter/MultiCopter rather than ArduPlane. There are many different versions, and people are always inventing new ones, too. Helicopter, Coaxial (two props, over and under) Tricopter, Quads, Hexa, Octa, Y frame Tri, Y frame with six motors, and many others. I am no expert, and it gets complicated, but here is the simple answer for quad vs hexa. If built for it, a quad can be less expensive to make and depending on the components, may be well suited for your needs. A hexa can be designed to carry more weight more efficiently, and can sometimes be made to recover from a motor failure. But the design is very important. Many n00bs (like me) run off and start "designing" monsters, with "upgraded" motors... but that, I believe at this stage of my self-education, is backwards thinking. Bigger motors (or more arms for the sake of having more) means more weight, more cost, and does not always mean more flight time. In a well built system, a hexa or octo will be more fault tolerant of a motor failure (but not always) and can carry more weight. But they are heavier, which means they break even faster when you hit the ground, and often (design-dependant) have similar flight times. The motors are designed to be efficient at different speeds, with different props. If you are building bigger just to build bigger, then you are (my opinion) falling into a classic trap. Built to your need, and you can build a more efficient system for your specific weight class. Which means less cost and more flight time for your purpose. But please bear in mind that most of my conclusions are the result of simulations and calculations. As I said, I'm new here too, and I have *maybe* eight hours of actual flight time :)
4) Generally what are the top speeds of these machines when you have a camera (goPro)?
I'm going to have to defer that question. I do not have a "standard" build. I just got a new 3DR frame, my first "kit" frame today, and it is nearly assembled already, but I have never flown it. My three other frames differ a lot, one is only 20" motor to motor, and the other two are 40" wide (but very very light; they are ... problematic ... so I have not share the design yet, still working out serious issues) but they have between 1.9 and 3+ thrust to weight ratios at 75% throttle. My calculations suggest that at 80%, my fastest configuration (specific payload - go pro would be ok, but not the case, battery configuration) I could achieve nearly 30mph vertical airspeed, but I do not trust the simulations (and my battery would last about 110 seconds) I'm cavalier with my experiments, but not that brave to have tried. For the planes, there are guys on here and on youtube that have videos w/ goPro at over 60-100 mph, and I think gas engine records for private RC planes might be in the 250+ MPH/Knots per hour. But gas engines are not great for autopilots, so I am lead to believe. Too much vibration. Also bad for photography for the same reason, in general.
5) Where get I purchase a fully built machine this month?
I have no experience with this. This technology and this community, my impression mind you, are more builders and designers and experimenters. But different people have different experiences, so some decide skip the electronics/hardware work and focus on software, some want to hack on the hardware, but not the APM itself, so they get it pre soldered. Some want to write and extend and tune software, some want to add cameras and get to photos quickly. Others just want to fly now. In the little time I have been here, this last person sometimes runs into problems. Because here, you can do anything, change anything, that means a lot of the technologies are more flexible than a pre-built standard, unchangeable commercial multicopter, pre fitted and packaged as a ready-to-fly (but do not change it) aircraft. The variations that can be had each allow for many opportunities to, lets say, make something that exceeds useful flight characteristics.
That being said, the people designing and building these full up ArduCopter kits are very experienced, far more than I, and they have a working design that many people use, and which I believe is used as a common development reference point.
Even still, I would expect that you may have to tune it for optimal flight (maybe even to avoid crashing by all but the best pilots.) I don't have one, so I cannot say. Ask me tomorrow :)
6) What will be the hardest challenge in learning to fly this machine.
As a fellow newbie, well, physics, it gets in the way all the time. Weight, motion, inertia, electrical shorts from swapped cables or bad soldering letting out the magic smoke... And the 3D aspect of motion. My transmitter has so many switches and sticks and buttons. In a panic, it is easy to push one the wrong way. And since I have no muscle memory, and very little time to react, I make mistakes. But that is once you are actually flying. Before you fly, you definitely want to tune your machine. But knowing how to tune it is hard, if you are a newbie. Knowing what it should do, and how it should act would be very helpful for a new flier, if only that were possible. If you have a prebuilt copter, and a good recommendation for firmware (or, better yet, an already tuned copter) then a brief bit of training, maybe a few hours, on what all the modes do, and some simulation or tandem stick trainer/buddy cable training would really make things much easier. When I first got started, while I was waiting for my DIYDrones stuff to arrive, I also got an AR.Drone Parrot to play with. The control scheme is very different, but it gave me a basic feel for what it is like to fly a quad, it is a lot cheaper, and it is fairly forgiving of mistakes and crashes. I think there may be a better option than AR.Drone someone else can recommend.
My apologies for writing so much, you helped me to organize some of my thoughts, and I needed a break from what I was doing. Happy hacking and good luck. Oh, and remember this, if you don't read anything else. As much as you want to just get to the flying part, tether, tether, tether (and hand-test.) That is, unless you are just here to fix broken things ;) And video record your early flights.
Simple mistakes (often) lead to spectacular crashes, even at no altitude or speed (don't they all end in no altidute and speed?) The video doesn't help prevent disaster, but it makes it a lot more fun for the rest of us (and you, much later) to see *wink* And it might help to know what went wrong for later.
Permalink Reply by Jason on September 13, 2011 at 1:52pm Mike,
Thank you for such a well thought response to my questions. Because my interest in this field is mostly the chance to add an exciting new service to our forensic arsenal I am not so interested in building it from the ground up, but that does sounds like a LOT of fun. More than happy to pay to have an experienced builder make the copter. I realize that even if it shows up exactly how I want I will need to learn and understand how everything works. Will also need to learn how to fix bad connections, replace broken servos, etc.
Where can I get the "flight simulation" software that will allow me to practice? Do these simulations allow you to control the simulations with the actual controls that will control your copter?

Permalink Reply by Michael Pursifull on September 13, 2011 at 2:58pm Jason,
I think it would be wise, if you want to explore the use of the technology for the purpose you describe, to put some money towards a local expert who has his own equipment, at least for some trials and exploratory work. Aside from the flight experience, you'll spend a lot less overall (no capital investment, and no repair costs) but there are other issues. When offering services, in the USA at least, there are some tricky areas with the FAA right now. If it is RC manual controlled remote video some interpretations are that it can be done legally w/o special approval and perform this as a commercial service. Others hold the opinion that so long as the commercial service is not principally to do with flight, but could be performed any other way, they can skirt the FAA issue, but both of those are very ... dubious ... and untested positions. Basically, in the US, at the moment, commercial application of UAVs require FAA COA, and they do not issue them. But that discussion is beyond my expertise ... and yours, but a third party can navigate the legalities on his own. All the UAV companies offer training for their gear, and they charge you for it, so I'm thinking there are legal options there, at least to examine the technology potential. In any case, that would be his problem, not yours. [I'm no lawyer, etc, no opinion on this legal stuff.]
As for the flight simulation, not only does it allow you to practice in the same way as you control the copter, if you get an APM, and a transmitter and a battery (or any good 5 volt power supply - I often use iPhone AC adaptors and just connect it to the servo rail ... or my Ham radio variable voltage power supply if I need more amps than just the receiver) you will actually use the transmitter, and the APM will fly the plane too!
It is called HIL or hardware in the loop simulation. At least two simulators work, the excellent X-Plane and the free FlightGear simulators. Both are nice bits of software, and at $29 X-Plane is a steal. I use both, because X-plane is very good but requires you to have a DVD and I do not always have an optical drive attached to my computer, and FlightGear is good enough for me at times, but sometimes I have technical glitches with it.
Read more about this here and here (both work with either plane or quad, or I have done quad with both), then go get an APM (kit if you solder, or pre-assembled if you do not). I do not get any kick-backs or anything, but for $300, you cannot find more fun IMHO. As I mentioned, you will need a transmitter/receiver of your choice ($50 - $500, depending on model - CyberCrash mentioned a great model for $60 that would be perfect for now, unless you have other RC hobbies already, in which case you'll already have one) and a 5 volt power supply (a USB cable and a recent model [green dot] iPhone charger will work, for example.
Come on in, the water's nice. But you can play without buying into the entire airframe, and get an expert for the business initiative to test it out first.
Permalink Reply by Jason on September 13, 2011 at 4:22pm The legal aspect of the flight is interesting, but the courts do not seem to care whether you had "permission" to capture data or not. they are not interested in whether or not you had a permit to operate the "cherry picker" when you took photos of the accident scene, Or trespassed onto someone's property to laser scan an accident scene. Now that does not mean the FAA or other agencies don't care and try and prevent you from doing your work.
In the 15 years that I have done this I can not tell you how many times opposing counsel "trespasses" on our client's private property to reenact an accident during closing hours or on city streets in the middle of the night while suing the city. Not once were counsel successful is getting work excluded because they did not have permission.
Your point about finding a local expert with experience to use the first few times is an excellent idea, my challenge is I do not live in a major city. How do I find the closest "expert'?

Permalink Reply by Michael Pursifull on September 13, 2011 at 5:05pm There are a surprising number of members of this community, from all over the world, and a review of some of the videos shows that a great number of them either live or spend a lot of time in rural areas right around the globe. Big cities are not great places to fly robots, I guess :)
Permalink Reply by CyberCrash on September 13, 2011 at 2:33pm Hey,
great Idea to join us. This machine is really amazing.
1) How do you control the camera remotely to take a picture when it is in the proper position?
Just with your cheap TH9x v.2 from hobbyking.com (60$). You got 8 Channels for that. Okay, 4 Channels for flying and one for all the functions (gps, alt hold, etc.)
2) Is there a way for you to see what it sees in real time?
It's called fpv. Look here for more info: http://www.rcgroups.com/video-piloting-fpv-rpv-469/
Visit hobbyking.com there go to Telemetry & FPV. Depending on where you live you get the 5,8Ghz or 1,2Ghz or ...
(for Europe 5,8Ghz. Get the 200mW transmitter + receiver for I guess around 70$)
Then you will need some Video goggles to watch your copter flying (from first person view): 320x240 video goggles about 77$.
3) 4 wings vs 6 wings what are the advantages?
4: cheap, works, stable
6: 50% more expensive on motors, ESCs and props. But with that new algorithm: it's more stable in case of prop-loss or motor failure. And it can lift more heavy loads.
4) Generally what are the top speeds of these machines when you have a camera (goPro)?
I did 65km/h once on just 50m but there is more possible, but not advisable.
5) Where get I purchase a fully built machine this month?
look on: http://store.jdrones.com/product_p/ackit1.htm or http://www.udrones.com/category_s/25.htm
6) What will be the hardest challenge in learning to fly this machine.
Reading the manual and following all the instructions on the homepage and especially the WIKI !!!!
And: !!!!
Before first lift of. Hold your copter with your hand from beneath and try to feel if the copter wants to move strangely. If so you probably did a mistace with the spinning direction of your motors or the props.
Good look.
Best to read through as many posts as you can so that you can avoid many mistakes some of us did and learned from.
PS: Build it yourself. The next crash will come and if you don't understand where everything goes you will not be able to fix it.
CyberCrash

Permalink Reply by Michael Pursifull on September 13, 2011 at 3:09pm Great advice. Since my experiences up to now have been custom builds from ground up, I put out too much of a caution and complexity vibe. I'm glad to have actual experienced folks provide a more balanced view. Well, at least that was my rationalization when I bought the 3DR frame to balance out my views :) really I just wanted more.... so be careful, Jason, this stuff is fun.
Permalink Reply by Jason on September 13, 2011 at 4:27pm CyberCrash,
Great stuff already I can see myself getting sucked into this world (in a good way). I agree with the build your own idea because even if I buy a pre-built I would still have to learn everything about it so when it does break (and it will) I can fix it. Heck I can see having two so if something goes wrong with one, we would have a backup. Nothing worse than being at a sight and your equipment fails with the client watching. Happen to me one time where I had to pretend the equipment was working (the client did not know any better) and come back later the day and do the work for real.
Permalink Reply by Patric Millar on September 13, 2011 at 9:11pm G'day All from the Land Downunder
First time posting here. Just want to say thanks to Jason for posting up. I was just about to start a very similar thread except I need a UAV to download data from a remote location. Now I can save everyone answering the same questions again.
Season Two of the Trust Time Trial (T3) Contest has now begun. The fourth round is an accuracy round for multicopters, which requires contestants to fly a cube. The deadline is April 14th.24 members
720 members
203 members
1355 members
207 members
© 2013 Created by Chris Anderson.
Powered by
