ABOUT ME:
I have always been a "hands-on" person with a passion for radio control modeling. When this contest was provided by NXP/Hackster, I just had to give it a try and see where my experience and challenges could be applied. I will admit, there were some ups and downs especially programming since I'm computer savvy but not a programmer. In my personal opinion about this contest, I have learned a lot from others whether asking questions in the group or researching blogs on the internet. Being a previous training instructor, I always had a knack for interfacing with people and to help those that need help or provide a probing solution to their problem.
FIRE FIGHTING DRONE:
The design I came up with using components provided by NXP was to include additional communications using a 4G network tied to an RPi 3 with extensive storage space for video and onboard programming using 256G Samsung Solid-State Hard Drive (SSD).
The intention (not yet complete - Challenge 2) is to include long-distance communications between the aerial drone and the user (fire tower) by a 4G network (Cell). The information would be more responsive and giving the required data to overlay video and current data such as drone/fire coordinance, heat signatures and continuously monitoring the fire threats and temperatures (use of PixyCam & sensor).
STAY TIME IN THE AIR:
My main challenge is the stay time in the air being a major factor. My current stay time is approximately 8 mins. I would have like to have changed the brand and size of motors to larger motors with larger propellers and battery packs. The idea of being a quad was a bit challenging with the hardware onboard due to weight distribution on the craft. If one were to make this into a commercial machine I would highly suggest (from a drone safety aspect) having 6 motors (Hexacopter) to help maintain stay time in the air. The idea of having additional motors is if an ESC or motor failure were to occur in the air allowing the drone to compensate for lost motor and would not down the craft.
J-LINKDuring the course of the build, some people had indicated they had issues with the J-LINK. My issue was trying to upload the BIN files to the FMU as the cable connections were not of high quality. In order for me to load the FMU programming, I did a continuity on all connectors and everything seemed to be in order.
It was by accident, I had squeezed the connector of the ribbon cable on the board and made a connection with the FMU - THEN - lost connection on data transfer when I released the connectors.
Apparently, the connector was defective not being able to reach the pin of the board OR when assembled the connector was not making good enough contact on the ribbon cable. I could not identify either.
The Fix:
Used Vice-grip pliers to lightly crimp and hold the connector in place to make connections. I made all of my uploads without error after the clamp was in place.
When the kit arrived, the FMU was not what I expected:1. Had to assemble the board to its enclosure.2. Do it yourself label installation on the FMU enclosure.3. The label identifying connection points on the FMU should have been darker – Very hard to read the connection points for your plug-ins.
FLYSKY FS-I6SI had issues with this radio losing communication between Tx and Rx while I was airborne during the maiden flight. I didn't have any stick control or RTL to land and none of my fail-safes I setup worked. I basically had to wait for the battery to slowly fizzle out and land. Thank goodness there wasn't enough wind that day during the flight only allowing the craft to drift approximately 25 yards from its take-off point.
The Fix:
I removed the receiver (FlySky FS-I6S) from the craft and installed my RadioLink A10. After the installation and setup of the programs to the RadioLink A10, I took the drone for another test flight. This flight all controls worked accordingly and flew great. However, the next challenge was my battery settings were incorrect and had a short second flight. Went into my setup and readjusted my parameters. The third flight seemed to be better for radio communications and parameter settings. I believe I need to do better calculations on my parameters for battery usage.
3D PRINTED CANOPY
All the widgets I have onboard this model I have seen some of my other drones open-faced without any protection. I thought with the RPi and other devices I need to protect them from the elements and decided to make a canopy. The objective of the canopy is to cover the onboard electronics in case of rain or debris such as dirt from landing or taking off from the ground not to mention "ASH" from forest fires.
Not having the skills working with Fusion 360 to obtain a Gcode for my 3D printing, I had to go to old school technology. Over the years I had built several fiberglassed parts for RC aircraft using Styrofoam, epoxy, glass cloth and fiberglass resin.
ADDED RETRACTS
One of the things I have found during my dealing with drones using cameras is looking at the landing gear. Which most drone pilots see if they don't have retractable landing gear. This project I figured I would purchase a set of Tarot retracts so the PixyCam could rotate for fires. WIth standard landing gear, this would block out what you are looking for "FIRES". After the installation of the retracts, I mounted on each leg HOLYBRO telemetry and Video transmitter. The antennas on both devices point inward during landing and take off but once in the air and the retracts go up the antennas point down separating the two and providing excellent transmission to ground control.
The modified landing gear helps provides:
1. Gives the drone a more streamlined and professional look in the air.
2. Allows aerial antennas for telemetry and video mounted to the retract legs.
a. When retracts raise both aerials pointed towards downward allowed excellent feedback to the ground station.
3. Allowing gimbal 360-degree rotation of Pixy camera without having to view fixed landing gear.
REPLACING PROPELLERSPropellers are everything to this flying drone.HoverGames HGDRONEK66 kit came with 9450 self-locking propellers. After a couple of flights, I decided to use my propellers from the 3DR Solo.
Even with the 3DR Solo propellers the motors never overheated and the performance was actually outstanding compared to the props that came with HGDRONEK66 kit.
Another noticeable difference when testing propellers, was the HGDRONEK66 kit propellers seemed to have more of a delay on takeoff compared to the 3DR Solo propellers. I believe the reason for this issue was the props were soft and flexed requiring you to ramp the motors up a bet higher on their RPMs to become airborne. The 3DR Solo props are little over 1/2" longer allowing more air for its pitch.
The material made for the 3DR Solo propellers is stiffer than the white props that came with the HGDRONEK66 kit.
Prop Comparison:
HGDRONEK66 Kit Propeller: (WHITE)Length: 9.4in / 239mmWidth: 1.250in / 31.69mm Pitch: 5”
3DR SOLO Propeller: (BLACK)Length: 10.0in / 254mmWidth: 1.219in / 32.24mmPitch: 4.5"
NOTE:
A compatible upgrade propeller is the Master Airscrew (MAS) known as the 3DR Solo Upgrade.
Purchased the 2 each Red and Green Night Flight LEDs and install them to the craft. Now having Nav lights for orientation especially during twilight and nighttime. The lights will help the observer identify the craft while in flight.
I always followed the rules of drone flying to include any RC Fixed-wing
WHEN performing the operation of the drone in a laboratory environment OR in the field all SAFETY INITIATIVES will be performed:
• SAFETY is a priority by recognizing and correcting unsafe conditions and behaviors.
• Comply with all Federal Aviation Administration standards and expectations and take action when standards are not met.
• Prior to flying, I normally contact my small airport nearby.
• Not fly above 400 feet· Stay within the line of sight.
• Not fly over crowds· Not fly over federal, state and local controlled areas, i.e., wildlife management areas, nuclear stations, military stations, community areas.
TESTING DRONE FOR FIRE (HoverGames Challenge 2)
After the month of September of 2019, Missouri weather "went to hell in a handbasket" quick! Either high winds, rain and thunderstorms and then winter set in with rain and snow. I never had an opportunity to do additional test flights due to inclement weather.
Owning 35 acres I have a pretty decent test area for my aerial and ground drones. The image below outlines the Ground Stations I have established and where I will have a fire to test the sensitivity of the Pixy and sensor.
I plan on performing a test at all three of the ground stations to detect fire in the field – and obtain a fire detection from the drone to the ground station to show the image of the fire and track it in flight. Since the drone has cell communications allowing it to travel at farther distances to help identify any fires.
The objective is for the drone is to send a collection of data (telemetry, temperature, and video) and collect all data to its hard drive at the same time while transmitting information back to the PIC or FT:
- Deliver the coordinates
- Stays within the area as an aerial monitor until a field team arrives. (pending battery life)
- Retrains copies of information onboard as well as information sent from the drones onboard 4G communications to the PIC or FT cell phone/tablet/laptop.
> Compass Issues
QGround Commander after connecting to the vehicle, Open Parameters-> Commander and change COM_ARM_MAG to 0.25. With the default, 0.18There are two compasses. One is a GPS compass and another one in the FMU. Disabled the FMU compassDisabled the cal_mag1_en) was to write to the drone the "nxp_fmuk66-v3_default.px4 (latest development build, for use with QGround Control)"
> Premature Low Battery Warning
Battery is a LIPO 11.1 volts when charged 12.53 volts. While flying I end up with the drone emitting a beep indication for Low Battery. I had to adjust parameters in MP for BATT_ to get the correct readings.
> Communication garble from PX4 to RPI1) Having to downloaded flight logs and submit it to NXP for evaluation to help identify the problem. This could help others that may have had the same issues with their equipment.a. Why failsafe did not work.b. Why there was a loss of radio signal and not being able to control the aircraft.
NOTE:
NO SUPPORT was provided from the experts on any of my requests to identify flight log radio issues.
2) Within its 3 minutes of flight time, I lost radio communications (FLYSKY unit) with the craft airborne.
Thankfully the craft stayed in one position and didn’t drift very far. The drone hovered while I tried several attempts to gain control again by moving elevation, yaw and pitch without success from my transmitter without any control to the airborne craft.
NOTE:
I replaced the FlySky radio with my personal radio system - (RadioLink AT10)
3) Remove the FlySky radio gear and use my RadioLink in its place to validate if it’s the FlySky radio is having issues or not.
a. The drone flew without an issue. Apparently, there was an issue with the FLYSKY radio system and not sure if it was receiving transmission properly due to the internal antenna or not. (my best guess)
NOTE:
Contacted NXP - Sent new FlySky radio system.
FUTURE PLANS:
Add onboard to the drone an N532 NFC NXP RFID Module. The RFID will be tied to the RPi 3 eliminating a manual push of the start button for drone startup by using an employee identification card for RFID or by fob.
Info/Blogs·
HoverGames Challenge; https://www.hovergames.com/
· NXP Semiconductors | Automotive | Security | IOT;https://www.nxp.com/
· Hackster.io; https://www.hackster.io
· PX4 Discussion Forum; https://px4.io/
· MAVLink; https://mavlink.io/en/· Raspberry Pi Foundation; https://www.raspberrypi.org/
· Master Airscrew 3DR Solo Propellers Upgrade 10X4.5; https://www.masterairscrew.com/products/3dr-solo-propellers-mr-sl-10x4-5-prop-set-x4-black-3dr-solo
Software
· Rapid-IOT-Studio; https://www.nxp.com/design/development-boards/rapid-prototyping/rapid-iot-studio-online-ide:RAPID-IOT-STUDIO
· EMLID Community;My Software Setup using 4G, Mission Planner, GStreamer;Drone Station 4G software
· PixyCam code for use with HoverGames NXPHovergames by igalloway; https://github.com/NXPHoverGames/PixyCam/commit/ce9175bfddf8b9023025da161ad283e3184490c8
· Thermal Camera and Sensor for use with NXP HoverGames drone kit KIT-HGDRONEK66 by Katrin Moritz: https://github.com/NXPHoverGames/ThermoCam
Electronics· SIXFAB; Raspberry PI 4G/LTE Shield kit; https://sixfab.com/product/raspberry-pi-4g-lte-shield-kit/· MikroElektronika (MIKROE); https://www.mikroe.com·
· Hexiwear Docking Station; https://www.mikroe.com/hexiwear-docking-station· SupTronics X825 SATA SSD Shield for Raspberry Pi·
4G (LTE) Telemetry and Control of Pixhawk using Pi3 by Artem Bulashev https://www.youtube.com/watch?v=iUctV-tqh_M
How To Do:
· Raspberry PI3 B+ How To Boot From An External USB Hard Drive Or Thumb Explained Tips And Tricks by Richard Lloyd
· Raspberry PI, Connecting Raspberry Pi to Pixhawk by Dennis Baldwin - https://www.youtube.com/watch?v=DGAB34fJQFc
3D Prints· HoverGames Frame Ring – by Norritt https://www.thingiverse.com/thing:3844850
· Pix 2 Pan Tilt by jhulberthttps://www.thingiverse.com/thing:3180802
· HoverGames Drone Kit HGDRONEK66 improved landing gear “T” by igalloway https://www.thingiverse.com/thing:3822544
· FPV/Drone battery tray (parameterized) by MacPara https://www.thingiverse.com/thing:1219149










Comments