Blogs

Hi all,

I would like to introduce you to an Open Source, low-cost digital FPV video transmission option!

https://www.tindie.com/products/FernTech/fpv-hat-for-raspberry-pi-/

Our FPV HAT for Raspberry Pi 2/3 has these key features:

• Includes open source software for driving RF module, incl. transmit, receive and test apps. https://github.com/FernRF/FPV-HAT

• Max. data rate approx. 500 kbit.

• 433 / 868 / 915 Mhz versions available.

• Supports AES-256 encryption.

• Useful for e.g. long-range and/or NLOS SD "analog quality" video links (e.g. 512x288@49fps or 640x360@30fps)

• OR high-res-but-low-frame rate DIY-surveillance/Robotics/RC/etc. applications

• OR anything else on a Raspberry Pi that could use a 500kbit wireless UHF one-way data-link!

• Set of two boards cost 70 USD (Rasberry Pi, antennas, camera etc. not included). In stock. Free Shipping.
  

The FPV HAT has been developed to enable wireless transmission of hardware encoded h.264 video streams from one Raspberry Pi to another.

The available bandwith is limited to 500 kbit which in turn limits the possible resolution while still maintaining reasonable fps and latency. Video quality is comparable to analog PAL/NTSC video.

But if e.g. your ground rover needs UHF ISM-band video in the 433 Mhz band in order to send the signal through walls or vegetation. Or you want to keep your video stream truly private using hard 256 bit AES encryption, then this might be the board you need!

Also, compared to analog video, bandwith use is very limited at either approx. 300 Khz (2GFSK/300kbit) or approx. 600 Khz (4GFSK/500kbit). User defined frequency offsets in software enables a number of concurrent streams within the respective ISM bands locally.

Please note that this system requires some knowledge of Linux and Raspberry Pi to use. Users may also need to adjust transmit power to comply with local regulations.

FernRF

To anyone considering a purchase from nemorc.com or dronada.com i would like to warn you of the following.

The store is owned by Phillippe Cosyn from Moldova, someone who is known to scam and has previously done so under the name RCMOLD-STORE.COM. (https://www.rcgroups.com/forums/showthread.php?1680460-RCMOLD-store)

He now operates under nemorc.com.

After transferring funds to him he stopped all communication. It was only later that i found out he is known to scam the RC community with other store names.

I hope this warning will save others from losing their hard earned money.

Beagleboard.org has released the BeagleBone Blue some days ago, a Linux-enabled computer for robotic applications.

The BeagleBone Blue is an open hardware platform that include a large array of sensors and IO; IMU, Baro, PPM for RC input, 8 channels PWM output to Servos, DC motor driver, analog input, PRU, GPIO, LEDs, buttons, serial, CAN, I2C, SPI, USB, WiFi, Bluethooth, 2-cell LiPo charger and wide input voltage range.

Jason Kridner has designed the BeagleBone Blue and he contacted me to implement the ArduPilot Flight Controller Software on BeagleBone Blue and make it a full blown Linux-enabled autopilot. Based on my previous experience with a similar project called the BBBmini, a BeagleBone Black or Green coupled with a sensor cape of my own, that has been successfully built and flown by numerous DIYers around the world, the BeagleBone Blue is natural extension of my design with the benefit of having a single board solution.

Experimenters can now have their copter, plane, rover, submarine being controlled with ArduPilot software powered by a versatile Linux platform with enough power to add cameras, additional sensors and use a large array of available robotic language like ROS, dronekit-python and Mission Control suites for just 80$.

https://github.com/mirkix/ardupilotblue

https://beagleboard.org/blue

I am not related to TI or beagleboard.org. BBBmini and BeagleBone Blue ArduPilot support are both pure hobby projects.

We've been told that the Basic Air Data, DIY oriented 8 mm probe performed well within a wind tunnel 

You can read the article about the wind tunnel here. Details within this pdf report. 

This is the second instalment in a series in which we test the precision of different drones and autopilots by flying them on a standard photogrammetry mission created in UgCS.

Last week we published a test of DJI Phantom 4 where we tested the precision of this drone on our photogrammetry route. In this week’s video we review the flight results from another popular drone autopilot from DJI – Naza-M V2. The Naza-M V2 is a popular autopilot from DJI used by hobbyists and professionals alike. In our case it was mounted on a DJI F450 frame.

We flew it on the exact same photogrammetry route as Phantom 4 with three different turn types – Stop and Turn, Bank turn and Adaptive Bank turn. As expected, with DJI Naza-M V2 being an older autopilot, it did not fly as precisely as the Phantom 4. Out of the three turn types the Adaptive Bank gave us the best results as far as keeping to survey lines goes.

Here is the video:

Download the KML files and test them yourself in Google Earth:
http://ugcs.com/files/autopilot_precision/DJI_Naza-M_V2_test_flight.zip

Get the newest version of UgCS here:

https://www.ugcs.com/en/page/download

Safe flights,
UgCS Team

This is my second post in the series called "FlightZoomer 3 Innovation Previews". The first was this one: Maximized terrain awareness for drone pilots

The innovation highlight presented in the video above is the synthetic camera, that offers an FPV experience without many of the disadvantages of a classic FPV approach. In the video 10 advantages of the synthetic camera are listed and demonstrated:

• No video data transmission needed -> less hardware, less complexity
• Truly worldwide range if there is nothing more than 2.5G cellular connectivity
• Highly improved reliability
• Cut the Gordian knot of zero cost Full HD FPV
• Principally no yello, no shaking
• No fog, no clouds, no twilight, never night
• Adjust the viewing angle during the flight
• Adjust the camera tilt and yaw angle during the flight
• Manual snap back view direction control (MSBVDC :-))
• Integrated optional zero lag gimbal

Of course there are limitations too:

• Close to the ground the synthetic view lacks accuracy.
• 3D cities are limited, 100% world wide coverage is only offered as aerial image laid upon an 3D elevation model (which offers still remarkably stunning views, especially when cruising a bit higher)
• The scenery stays static. You cant observe non-static objects, like traffic, cars,...

The solution is implemented using the MapControl of the UWP programming stack. It offers a comprehensive API to place a camera over the 3D landscape at any place and specify all the parameter, that we need (including pitch and bank angle). B.t.w. this is also the reason, why this solution is not easily portable from Windows to Android or IOS (in fact not doable at all): the Android or IOS map API do not offer the needed capabilities for this feature. Google does not support to place a camera at all at a particular position and altitude and Apple does not support the combination of setting pitch & bank while using 3D maps with an aerial image.

An important detail is the possibility, to additionally feed the video from a real FPV camera into the app. The real camera view can be placed on top of everything else as a small, moveable and resizable overlay. This feature perfectly complements the overall package, FlightZoomer offers.

The Ardupilot Development Team held an "un-conference" (aimed at developers) in Canberra, Australia last month. The conference included several talks, along with flying in the field.

Ardupilot developers present included  well-known Plane, Copter and Rover  leads Andrew Tridgell, Randy Mackay and Grant Morphett, and developers Paul Risenborough (EKF), Leonard Hall (control and auto-tune),  Philip Rowse (Pixhawk2),  Michael Oborne (Mission Planner).

For those interested, the first two talks and demos  videos are available on Ardupilot.org following these two links: 

Talk and demo by Andrew Tridgell, Talk and demo by Paul Risenborough.

I love the OpenMV board, a $65 open source computer vision module that combines a camera and powerful M7 processor with a nice built-in Micropython interpreter and OpenCV-like computer vision library. It's also got a fantastic IDE that makes using it the simplest computer vision experience I've ever tried.  You can see in the video below how I used the board by all by itself (it can also control servos) to drive an autonomous car to follow a track:

Because the board can be easily scripted in Python, it can be configured to be pretty much any computer vision sensor you want it to be, from optical flow and line detection to full-on object recognition and tracking. Even better, it can now instantiate as a MAVLink sensor, so you can add it to any MAVLink-compatible autopilot as a sensor. 

Here's the code for making it a MAVlink optical flow sensor. 

And here's the code for making it a MAVLink April Tag sensor. April tags are the best way to recognize codes from afar with computer vision. Demo below and more info here:

Hi Guys!

This year we will be at XPONENTIAL 2017 (May 8-11, Dallas). We have a small booth for startups. We are very excited because it is the first fair we attend in the United States.

Cool Atomo prototyping kits from HK.

"There are a number modules, including I/O boards, network expansion boards and even power supplies for bigger projects. For example, the project below has four I/O adapters and a power supply all connected to a Raspberry Pi. This means you can do some really interesting things with robotics and even hydroponics with the package the size of a hunk of cheese.

The kit costs $39 for early birds and should ship in June. You’ll be able to buy more packages and mix and match them as necessary."

From TechCrunch

From The Verge:

The official Raspberry Pi magazine just announced that over 12.5 million of the affordable little Linux boards have been sold since the original Pi was launched in 2012. As The MagPipoints out, this puts the Raspberry Pi past Commodore 64 sales, according to some estimates. That would make the Pi the third best-selling "general purpose computer" ever, behind Apple Macintosh and Microsoft Windows PCs. As commenters have pointed out, this isn't a precisely fair comparison, because there were other Commodore models than just the 64, but it's still a nice milestone all the same.

It's interesting that the most sold is the latest, the RaspberryPi 3, which is in my mind the first one that really reached the potential of a single-board Linux computer (built-in Wifi and decent processing power). So that bodes well for continued accelerating growth.

By comparison, as of four years ago, the Arduino team had estimated that there were 1.4m Arduinos and clones sold and presumably it's several times that now (Arduino was launched in 2007, five years earlier than Raspberry Pi).  The two don't really compete -- Arduino is best for physical computing and interfacing with sensor, while RPi is best for video and running Python code fast -- but I do find that I almost never use Arduinos anymore. There are enough good RPi "hats" that can read and control other hardware that I rarely find the need to add an Arduino into projects now.

We long ago decided that the controller of the Hybrix20 was going to be a Pixhawk. Some customers tell us that Pixhawk is not the right platform for our product but we would like to work on those few things that are missing to fit a drone of these features.

The first step has been to ruggerize the enclosure and the connectors. The second step is to adapt the software to the special requirements of Hybrix20. 

While we thought how to do this I would like to show you the box we have built

Jose Luis Cortes

Quaternium

Nice to see a 3DR Iris+ there, but I'm slight afraid it got totally destroyed afterward ;-) Check out DroneClash for more mayhem. Looks fun!

In a real life video game of sorts, four teams will battle simultaneously. They can use as many drones as they like, but each team is only allowed one FPV video stream to their drones. This means, in practice, only one drone can be tele-operated per team. But, teams may switch between drones or create autonomous drones, and anything in between. As long as it flies.

Precis RTK system is being used for some pretty cool stuff. Mr. Stephen Noble, owner of Drone Surveying, used Precis RTK boards to provide land survey and aerial photography service in Australia. By leveraging L1/L2 GPS frequencies for accuracy and reliability, when fixed in RTK or PPK mode, their equipment provides centimeter absolute accuracy. Check it out.

Contact Us, click here. 

Here at SPH Engineering we have a fleet of various drones we use for diverse research and development projects as well as testing new releases of UgCS.

When it comes to using drones for land surveying, mapping or 3D reconstructions, flight precision plays an important role. This is why we decided to test the precision of different drones to see how well they compare to one another. For that we prepared a short photogrammetry mission in UgCS to see how well does the actual flight trajectory correspond to survey lines.

In the first part we tested the DJI Phantom 4 using different turn types: first flight with Stop and Turn, second - Adaptive Bank. See the flight comparison in the video below:

 

We will try and publish one drone test per week. Next-up will be the DJI Naza-M V2.

Below are the KML files from the test route and the flight paths. You can import and view them in Google Earth.

Test Route
Phantom 4 Adaptive Bank Turn
Phantom 4 Stop and Turn

Get the newest version of UgCS here: www.ugcs.com

Safe flights,
UgCS Team

Hi, this is my first DRONE project. I make a cheap GPS tracker for DRONE. Here are some pictures, any questions you can ask me :)

Protocol: UART

Speed: 9600 bit/s

There's much to like in the new Beaglebone Blue board. Along with the usual Beaglebone processing power and Linux support, it has built-in IMU sensors, good WiFi, proper connectors and (for rovers) motor drivers and wheel encoders. It's also just $79, and supports ArduPilot out of the box!

But before you toss your RasperryPi 3/Navio 2 combo (which is awesome but costs more than twice as much), here are a few things to keep in mind. 

• No built-in GPS (not a big deal, since outboard GPS is the standard)
• Not really designed for video. There's no HDMI in or out, and no GPU. You connect a USB camera but it's not as well supported as video is on the RaspberryPi 3
• Community support for Beaglebone is not as strong as RaspberryPi

That said, for ground rovers I think it's a very interesting choice because of the built-in motor driver and encoders. I'm going to give one a shot. 

Schulman points out that while technology features such as geofencing offer safety options, they aren’t designed to be used as a sledgehammer solution.  “The notion that airports and drones never mix is an oversimplification,” he points out.  “We have many customers doing important work at airports, enhancing the safety of the national airspace system. Similarly, our live geofencing system can help prevent drones from entering wildfire locations, but we also know that firefighters are using our drones to keep themselves safe and to help fight fires. Completely disabling a drone in such locations would result in a net detriment to public safety.”

“Here is the lesson we have learned that only comes with actual operational experience across hundreds of thousands of customers: while geofencing is a great feature that helps prevent inadvertent operations, it will always require a balanced approach involving exceptions. Requiring drones to simply turn off when they are near airports is not the right solution to safety concerns.”

Schulman’s testimony points out that identifying a single technology in rulemaking is a mistake, locking the industry into a soon to be outdated solution.  “…locking in any specific technology mandate will discourage DJI and our colleagues in the industry from continuing to rapidly develop new safety technologies,” he writes.   “A requirement to implement the best technology available today discourages manufacturers from developing the even better safety technologies of tomorrow.”

sUAV technology is now on a plateau, anyone could build a drone to bid a contract, and fooling customer with flight time, or lowering price by incapable hardware. Lily fails, Parrot is cutting jobs, This market is NOT healthy.

Even worse, an average user still crash their DJIs a lot: flying a dji from balcony then flip vertically to the ground, racing drones still use analog modulation to acquire flickering video and crash into trees, false GPS lock for careless pilot and then fly away...  although they are all “pilot error”.

Where to go to have fun? Depends how serious you are.

Now tesla has autopilot, mercedes has E class also doing L2 autopilot, but a $5000+ sedan is way too much comparing a DJI toy, or any rc toy including a turbine jet. Or you can join Udacity course to have a nano degree on this. Well, it depends how serious you want to go, and last but not least, put your ass in a car with a RESET button is not the  flying a drone with RTL switch.

At least you can get a L2 autopilot kit to drive a honda with no hands. Take a look at http://comma.ai

But, the elixir for the future, really is deep learning?

Average model hobbyist don’t know what computer architecture is, nor what difference between linux kernel and distro. They want to FPV, thus, they will not have “fun of the future”.

YOU WILL NEVER KNOW UNTIL YOU HAVE DONE IT, RIGHT?

“For the future” you will need at least:

• ROS capable SOC platform, a raspberry pi 3 is minimal for CPU, jetson from nVidia is de-facto

• Lidar or TOF camera for depth point cloud, Hokuyo LIDAR, intel RealSense, or old Hokuyo LIDAR, or Garmine Lidar lite with a motor.

• Machine Learning coding/training experience: depends how nerdy you are. Python based platform is popular, you don’t need to worry about stack overflow at least.

A good starting point will be: https://github.com/DJTobias/Cherry-Autonomous-Racecar

Putting this bunch of $2K hardware for a starter, it’s better to roll on wheels on the ground, and makes more sense for developing initial obstacle avoidance algorithm. Well, driving a car to avoid crash by itself, equals crash it purposely, is the future.

Now the stake has been raised. Are you ready?