What would you like to see in a thermal pan tilt platform to make it effective for UAV usage?
I am with a company called SPI Corp and we manufacture a low cost multi sensor pan tilt platform called the M1-D. The M1-D initially was developed for use on ground vehicles but due to its light weight, small size and low cost it has become very popular with folks in the aerial imaging world including both manned and unmanned craft. We are anxious to make this an even better product for UAV usage especially in the low cost market. We are reaching out to the UAV community for help in this endeavor (especially you guys at DIY Drones). What we would like to know is what technical features would make our system even better for use in UAV's? If you could have anything you want in a low cost UAV Gimbal imaging system what would you ask for? What would be the bare minimum you would need to make the M1-D pan tilt thermal camera a workable UAV solution? Current imaging systems are extremely expensive. It is our hope that by starting with our current low cost solution, and with a little help from the community, we can come up with a low cost UAV solution that will enable everyone to incorporate advanced imaging into their UAV project.
Our current system specs:
Fully enclosed pan tilt gimbal
Sensor Payload - Thermal Imaging, CCTV and Laser Pointer
Diameter of ball - 4.5"
Diameter of base - 5"
Height - 6.5"
Weight ~ 2lbs.
Pan - 360 degree continuous
Tilt - 90 degree
Pan Tilt Speed - Pan: 0.05~220°/sec; Tilt: 0.03~140°/sec
Video Output - Analog Composite single line NTSC or PAL
Communication Camera Control - RS/485 PELCO-D protocol
Inverted Mounting - Yes capable
Hardware specs such as size and weight are obviously difficult to change. What are the key features that you would need to integrate into your UAV project? Are there specific interfaces that are commonly used? Please post comments and let us know what you would like to see in the M1-D to make it even better for UAV imaging.
We look forward to your input.
Thank you
Randall Kolchins
SPI CORP
Comments
Probably if you guys create just a small thermal cam with a "cubic" shielded metallic enclosure and a pin-out containing GND, VCC and Video_OUT you would start to sell them to many of us. We're used to create our own cheap pan/tilt devices. Adapting it with another cam side-by-side would not be that hard.
Another idea would be create a modular system:
1 - a cubic enclosure camera (basic pins: power and video output);
2 - arduino compatible pwm control board (available serial control: power on/off, pan, tilt and zooming);
3 - pwm tilt/roll pod (two servos);
4 - pwm zooming lens (one servo).
Unfortunately, your system is too much big and weight. Quark is the core I think would provide the best FLIR solution for UAV. It has just 17 x 22 x 22 mm w/o lens and it weights just 23g to 28g depending on the lens: http://www.flir.com/cvs/cores/view/?id=51266
Boscam HD19 is not thermal. However it is a good example of a really interesting almost "cubic" enclosure form factor like I said above.
http://www.boscam.cn/productshow.asp?id=19
Would we really need the laser?
Mark Lanning
Interesting thought. So if we provided a camera pack that included the thermal, visual and laser you would be interested? What if it were a simple box that could mount into a gimbal like the Z15? What would you need for an interface to get the video out and control the camera functions? is continuous 360 rotation important because without a slip ring you would be limited? This may not matter though for your purpose. Let us know how you think it should work.
I think there is actually a lot that you can do. Strip out the pan and tilt mechanisms for one.. provide the sysem in a simple camera mount sort of thing and allow UAV operaters to use existing expensive gimbals they have already purchased.. I don't think you can compete against the large array of companies making gimbals and often these gimbals are designed for the frame that they are mounted to not to mention the flight computer being used. I would focus on the FLIR and camera system more than the pan/tilt/roll and leave that up to companies like DJI and their Zenmuse Z15 gimbal and others like that.
I would much rather have a camera unit I can swap out while keeping my gimbal fine tuned for the frame.
Mark Lanning
We have had other people mention tilt and roll. I am not sure how we would do this however. Maybe if the unit was nose mounted pointing forward? Yes the system is too heavy for smaller UAV's but there is not much we can do about the weight (we are trying though with some material replacements). We have some choices of optics on the thermal side with 18 degree to 24 degree FOV plus up to 8x digital zoom depending on sensor resolution. We will really look at the tilt roll issue going forward but I think it would take a whole new design. We would love to show you and everyone the unit. We are working on some demonstration videos right now. I will keep everyone posted on that.
Many drones have a "FPV" system using a TX/RX that takes a analog composite input. So you're current output should integrate easily. It is possible to use RS/485 protocol to communicate to the apm but a TTL 5v serial protocol or PWM would be much easier to use as we would not need to change to -7v,+12v logic levels.
I would recommend keeping the control of the pan tilt as simple as possible and leave it too the apm control.
Absolutely Randall. A lot of thermal imaging devices only have digital zoom and no optical zoom. Low cost is ALWAYS relative! I guess a print out of the signal analysis of the operation of the camera would only show what libraries would have to be written on the APM. But that's the fun of DIY!
It would be great to see a higher resolution visible camera system, including options for both analog (NTSC/PAL) and digital output (this would be nice on the IR cam too) (says something about this on the datasheet, but really not clear at all). This would make it easier to use in computer vision systems, and improve compatibility with digital video links. Also, it would probably make it far easier to capture still images. Weight and cost definitely seem reasonable. One important thing would be the dynamic capabilities of the system. In the control of the system, does it provide good feedback on status fast enough to do closed-loop control for stabilization? How quickly can the commands be updated? 1Hz? 50Hz? 500Hz? It would be great to see additional interface capabilities beyond RS-422, such as CAN or possibly USB.
Just some thoughts...
Well i just wrote this and it didn't post. I will say it again. It should be cheaper than the 3900 if you don't need the Coontrol station that goes with it. send a couple of these out to us and let us figure out how to make them work. Using a computer and joystick. Just sayin
-
7
-
8
-
9
-
10
-
11
of 11 Next