Our SAR team runs the State of Florida and we are in the process of adding drones to our deployment list. We have chosen the X8 because of it being a stable platform. but we are unsure as to Motor/propeller combination. Auto Pilot/GPS/UAV system and video. I have read that many of you use for similar applications like power line inspections and such. Can you offer some advice? Our goal is to fly from bas camp to the desired search area, put on auto nav and search the area with the on board cam system. We would like about 1 hour flight time.
Any help would be greatly appreciated!
take a look at http://stefan.gofferje.net/projects/uav-projects/x8-setup-list where Stefan collects data of X8 configurations. I'm using a Hacker A30 10XL with a 12x8.5 prop. Using 10000mAh 4S packs you can easily get more than an hour of flight time.
The other factors hugely depend on your budget. On the low cost end APM now is one of the best systems I have flown yet, comparable and even surpassing more expensive solutions when properly used. When you are willing to spend more there is an enormous variety of systems available which you will have to compare for yourself as the best combination of features and expense depends on your specific requirements.
Nicolas--thank you very much. What do you recommend for a video system?
If you aiming for really long flight`s, you could use 3g/4g GSM to transfer the Video stream and telemetry.
Thanks Bernt--where is that software at on the internet?
That again depends on what you want and what you can afford. I would not really recommend 3G; I have the feeling that most of the work using cellular technologies currently is experimental.
Cheap 2.4GHz systems are capable of very long range, too. Look at the FPV flyers who are able to achieve >30km range with analogue video systems. One of these, in combination with a tracking antenna, would be my recommendation. Additionally, you are independant of reception problems.
Hi Michael, a few things:
At this point in the technology development SAR searching via a real time video downlink isn't really feasible. The problem is that you're limited to standard definition video which makes spotting something as small as a human figure difficult (vehicles or downed aircraft have a better chance of being seen). The exception to this could be FLIR where you could come across a heat signature where there shouldn't be one.
The way around this is to use an HD video cam or a large pixel count still camera taking images at close intervals. They would have to be reviewed upon the aircraft's return to the ICP. But this absolutely can work.
Australia is much further ahead than the US in terms of using drones for SAR. They have been running an annual "Outback Challenge" for a few years now where teams compete to locate a human sized manikin located somewhere within a given area. I believe last year's winners developed some image recognition software that they've since released to the public. Very impressive stuff.
At this point an APM is almost certainly the best flight controller for SAR due to its autonomous flight capabilities. I don't know how much you've experimented with the Mission Planner Software yet but it fits well with SAR needs. You can pull up an area of interest in Mission Planner, click at multiple points to draw a polygon around it, tell it what kind of overlap coverage you want and how high you want to fly, and it will automatically generate all the waypoints needed to cover the area, ready for loading into the aircraft. Very damned impressive!
The X8 is a good aircraft choice as it's established. A potential newcomer is the Techpod, capable of 2 to 3 hour flights. It's more of a slow, gentle flyer with a very large equipment bay. I'm experimenting with a Techpod.
A final concern would be the control and video RF links to the aircraft. You mention a hour flight time but also of importance is how far the aircraft will be from its control point. You generally need to maintain a line of sight to the aircraft, although this requirement can be lessened a bit if using lower frequency RF links. Fortunately Florida is pretty flat and you could concoct some sort of extensible mast to raise your antennas. It would probably give you many miles of range.
You have described my program exactly! Many in the USA have no plans to use drones for SAR but I truly believe it will be standard equipment in the next 5 years or so. We do want to have a FLIR package in the drone but as you probably know those are incredibly expensive but probably the best sensor for the job. In the past we had the Sheriffs Helicopter for a FLIR platform but due to the slow economy they have been greatly cut back in their usage, hence the need to look at drones.
I believe that Garmin is coming out with a new line of cameras soon that might work out better than Go Pro. But you have described the mission exactly. Take good footage of remote areas of interest and let the base camp review for possible finds. Then we can task teams to go out and take a look. But we are also looking at using Quadcopters to do detail work in hard to reach areas that we work in --like the Everglades!!
As for radio feeds we are thinking of a Balloon to raise up the antenna--has anyone tried that before? In general the drone would not be more than 5 miles from base at any one time. So LONG fly away's are not in the current mission model. We are more concerned that the drone can loiter for a good period of time and take lots of photos for review.
Do you have a recommendation for software and auto pilot?
Michael, FLIR does work fairly well...some of the time. But I know of a lot of big misses. In a warm environment (Florida, cough, cough...) it's very hard to see a warm body. But there's something about FLIR than enraptures SAR management so selling a UAV becomes a lot easier when you can say "FLIR capable". I think the really big deal will be automatic computer image processing like what's being done in the Outback Challenge.
Since it sounds like you're thinking along the lines of doing image analysis after return to the ICP, I agree you can do a lot better than a GoPro. The Sony NEX-5 is very popular for high resolution mapping for imaging and I think it would work very well for SAR.
In terms of software or autopilot I'm not up to speed on the expensive, really exotic stuff. For normal folks the only two real options are either APM or DJI. The drawback to DJI is its ability to do preprogrammed missions is very limited.
For this APM excels. I already mentioned its ability to automatically generate a search grid with waypoints. And it it loses the control signal it can be programmed to automatically return to its launch site and land itself. That's pretty damn neat!
To do it you need the APM 2.6 from 3D Robotics and the GPS module. All the rest of the software for it is open source and free. You can download Mission Planner anytime you want and play around with its planning functions and set waypoints. I will concede there's a pretty considerable learning curve to it all but it's well worth it.
Given the flatness of the Everglades and a limit of 5 miles a balloon could work but it could be problematic. You really don't want to run a lot of coax cable between the transmitter and the antenna.
Possibly a better option would be to use a long range UHF control setup like EZUHF by ImmersionRC. It's a transmitter/receiver combo that piggybacks on to an existing RC transmitter but instead transmits around the 430 MHz range. That frequency will actually punch through vegetation fairly well. A 30' mast might get you high enough so that the RF transit of the vegetation is minimal and everything would work OK.
For FPV you could use a video transmitter in the 900 MHz, which will sort of make it through vegetation. But the important signal is the control signal.
thanks Tom--we are getting the APM 2.6 and looking into antenna arrays for the control signals....
do you have any leads on the imaging recognition software they were using in the Outback Challenge? I have been trying to track them down with no luck so far?
Yep, that's the one!