Blogs

Thanks to our friends at Vice Magazine/Motherboard TV, who did the excellent Drone On documentary on us and the domestic drone movement, I'm going to be doing an Ask Me Anything (AMA) on Reddit today at 12:30-3:00pm PST. These things can get pretty wild, so I'm warming up my typing fingers already!

From Vice:

Chris Anderson has always wanted to take to the massive social-news site to field questions from hobby droners, makers, and curious long-tailers alike. But his job as editor-in-chief at Wired, a Condé Nast publication, always seemed to get in the way. Not like he couldn't ever have done an AMA on his own volition, or anything. Only seeing as Reddit not only flew under Conde's wing from 2006-2011, but even shared the same office with Wired magazine, Anderson tells me the thought of being asked anything on Reddit, of all places, always felt just a little too incestuous. Until now.

Anderson, of course, recently left his post at Wired after 11 years, a tenure that saw the pop-tech monthly winning eight National Magazine Awards. Why depart? In a word, drones--or, more specifically, 3D Robotics, the manufacturing spinoff of Anderson's fast-growing, nonprofit online hobbyist drone hub DIYDrones. Anderson has decided to throw all his energy behind the company, which to date has sold more than 10,000 custom auto-pilot kits. (That's 10,000 otherwise traditional RC planes now capable of flying autonomously, a fleet far larger than that currently deployed by the US military.) It's early yet, but consider the initial figure: 3 DR, already a multimillion dollar company, just raised a VC round worth $5 million.

"We're going through significant expansion," Anderson told me. "We've ramped up the Tijuana facility [satellite to 3D's base in San Diego], and we're opening a 3D Robotics North, in Berkeley."

That said, the time has finally come. In light of Anderson's 180, and both the recent release of our latest documentary, Drone On, which features 3D Robotics, and Anderson's new book, Makers: The New Industrial Revolution, the man behind "free," the "long tail" and the democratization of unmanned aerial vehicles will open the floor to questions tomorrow, Wednesday, Dec. 12, beginning 3:30pm EST, at Reddit's AMA page.

Do tune in, drone out, and ask away. The wait is over.

While looking around for a 433 tracking antenna we found out that:

1) There isnt much choice for 433Mhz patch antenna 

2) There isnt much choice for a solid housing

So we decided to custom make one ourselves. It has a 433 patch antenna that fits onto RP-SMA of the 3DR radio. The radio sits inside the housing and the USB cable comes out from underneath it. The metal housing has 3 servos and a old ardupilot. It takes 5V and be powered by a 2S or 3S battery. The base of the tracker fits into a standard tripod mount. We tested it out and the range is much longer than standard 3DR radio with antenna and it works with the mission planner. We also notice the plane position on mission planner updates more frequently. 

From Gizmag:

For some time now, engineers have been researching the possibility of equipping UAVs (unmanned air vehicles) with robotic arms. This week, DARPA announced that a successful test of just such an “armed” UAV has been performed.

DARPA utilized an existing V-Bat UAV, made by California-based MLB Company. The stock V-Bat can fly horizontally to quickly reach its destination, then hover in a tail-down orientation once it gets there.

The DARPA team added a folding robotic arm to the UAV, along with an inexpensive stereoscopic computer vision system – this incorporated one camera on the end of the arm, and another on the aircraft itself.

For the test, a 1-pound (454-gram) payload was attached to a hook on the end of the arm. Using its vision and GPS systems, the V-Bat then autonomously identified a step ladder, which was the object on which it had to place its payload. It proceeded to maneuver itself up alongside the ladder, extended its arm by six feet (1.8 meters), and placed the payload on the ladder.

“Our goal with the UAV payload emplacement demonstration was to show we could quickly develop and integrate the right technology to make this work,” said DARPA program manager Dan Pratt. “The success of the demonstration further enables the capabilities of future autonomous aerial vehicles.”

A video of the test (which apparently took place last September) can be seen below.

From the ROS blog:

The Robotics Institute at CMU has been developing systems to learn from humans. Using a Machine Learning class of techniques called Imitation Learning the group has developed AI software for a small commercially available off-the-shelf ARdrone to autonomously fly through the dense trees for over 3.4 km in experimental runs. They are also developing methods to do longer range planning with such purely vision-guided UAVs. Such technology has a lot of potential impact for surveillance, search and rescue and allowing UAVs to safely share airspace with manned airspace. 
Watch the flight in unstructured environments below.

Warning: this is a political video from an animal rights group, so it's very anti-hunting. However, the analysis of the damage to the hexacopter when it was shot is interesting. 

Silly political gestures (banning model airplanes with cameras!) are par for the course in Berkeley, where I live and the 3D Robotics Bay Area offices are located. Thankfully, they usually don't pass or aren't enforced.

From Patch. 

Even as the Alameda County Sheriff’s Department considers buying an unmanned aerial vehicle, Berkeley City Council will discuss declaring the city a No Drone Zone. 

Arguing that drones are unsafe and pose a threat to civil liberties, Berkeley’s Peace and Justice Commission is recommending that the city council adopt a resolution this month proclaiming: 

"1. Berkeley a No Drone Zone and instructing the City Attorney to perform the necessary legal tasks to transform this declaration of a No Drone Zone into an Ordinance for the City of Berkeley wherein drones are hereby banned from airspace. 

2. That drones will not be purchased, leased, borrowed, tested or otherwise used by any agency of the City of Berkeley over the City of Berkeley, including drones in transit.

3. That exemptions will be made for hobbyists to continue to fly remote controlled model aircraft in specified areas, away from dwellings and the urban cityscape of people and buildings as long as those devices are not equipped with any kind of camera or audio surveillance equipment."

It’s not clear what the “legal tasks” establishing Berkeley as a No Drone Zone would be, of if such an ordinance could be enforced. While the Federal Aviation Administration is still drafting the rules that would regulate what the agency calls unmanned aircraft systems, local law enforcement agencies and other institutions across the country are taking steps to acquire and deploy drones. 

Earlier this month, Alameda County supervisors opted to examine the merits of a cop-operated drone in more depth before approving the purchase of the $31,000 laptop-sized flying machine the sheriff said he needs.

 

(image from No Drones Illinois)

From Phys.org:

A team in Japan are hoping pedal-power will beat the world record for a human-propelled plane—in a flying machine made from polystyrene, they said Monday.

Team Aeroscepsy say a professional mountain biker will pilot their "Gokurakutombo", which has a wingspan half that of a Boeing 747 jumbo jet, in a voyage they hope will clock up 120 kilometres (75 miles). "We developed a new plane from scratch to break the world record," said Shinsuke Yano, who heads a collection of 10 engineers and enthusiasts. The Gokurakutombo—a phrase that means "Happy-Go-Lucky" and is also a pun on "happy dragonfly"—has wings that measure 35.6 metres (117 feet) tip to tip, but weighs just 37 kilogrammes (81 pounds) thanks to its polystyrene foam and carbon fibre construction. Yano said the world record for human-powered flight was set by a US university in 1988 with its pedal-driven Daedalus, a contraption named after the father of Icarus in Greek mythology. The Daedalus plopped into the Mediterranean just a few metres short of the Greek island of Santorini after flying 115 kilometres from Crete.

Team Aeroscepsy, who all work at Japanese motorcycle-maker Yamaha, hope to launch their flight from a field at the foot of Mount Fuji, and head for the Pacific Ocean next spring. Their contraption can take advantage of thermal air currents but does not need them to launch and fly. "Because it's a long flight over four hours and you need to keep pedalling with power that is required for climbing uphill constantly, you need a professional cyclist," Yano said. "We are pretty confident about reaching a new record. We know from past tests that our aircraft has that capacity. The most difficult part is reading weather conditions. Light wind can upset the fragile plane." (c) 2012 AFP           

Read more at: http://phys.org/news/2012-12-japan-power-aims-human-flight.html#

First, load the fuselage to an inability to move. And then begin to glue stringer requiring moving the fuselage.

Hey everyone,

 I am happy to announce that aluminum molds for mass producing the techpod are nearly ready to start cranking out airplanes!

Thanks to everyone who helped me out on kickstarter! You can still pre-order your own techpod today at hobbyuav.com They will be ready to ship in January.

Slashdot is carrying a story about new research published by the good folks at Carnegie Mellon and Coherent Navigation detailing a series of classes of GPS attacks beyond those previously demonstrated. 

Of particular note to this community, the research details vulnerabilities found in the uBlox 6H GPS receiver. 

The researchers go beyond the typical jamming and spoofing techniques, and also examine weaknesses in GPS messaging and in attacks against the embedded operating systems and applications which communicate with the GPS modules. 

The uBlox 6H does not fair as badly as some of the other modules, and besides spoofing, the only other identified issues is described as "Vulnerable Week #"

"Vulnerable Week Number Date calculations in GPS receivers are done using the Z-count, which consists of a 10 bit Week Number (WN) and 9 bit Time of Week field.

Here is a compilation of clips from the best flights from my first month flying my new FPV Deep Reaper wing. The first minute or so is my maiden flight, and the rest of the video features clips from various flights around the Denver area.

Here's a shot of my plane and ground station:

And here's a close-up of the plane:

This plane turned out a bit heavy for a Deep Reaper (around 6.5 lbs) so I am currently working on finding the right motor combination to keep the plane as light and balanced as possible while having enough power, speed, and endurance. Flight times are currently around 25 minutes with an NTM 35-36 motor and 9x6 prop. Next I plan on trying the lighter NTM 35-30 motor with the same prop to see if I can cut a few ounces and get the correct CG without putting led weights on the nose.

I am using an 800mw 1.2 Ghz video TX with a cloverleaf antenna on the plane and crosshair antenna on the ground. Control radio is a ChainLinkDare UHF LRS mounted on a cheapo FlySky TX.

F-117 BUG(Butterfly Unmanned Glider) Project Part 2:

Url to Part 1:

http://diydrones.com/profiles/blogs/f-117-hybrid-glider-project

I hope you enjoy my video everyone!  Let me know if you have any suggestions. 

Things to do:

1. Buy 4 screws with clip holes drilled down the shaft
2. 4 screw clips and 4 screw bolts
3. cut wing and paint wing black
4. drill holes in Styrofoam plane for screw pins
5. Test Design and Build CAD Design for Video Part 3 

From Darwin Aerospace, the "Burrito Bomber"

Earlier this year, the world was brought to its knees when it was cruelly teased with the promise of tasty Mexican treats being delivered via quadcopter. We believe everyone deserves carne asada when they want it and so today, we make that dream a reality. We’re proud to introduce: Burrito Bomber -- truly the world's first airborne mexican food delivery system.

It works like this:

1. You connect to the Burrito Bomber web-app and order a burrito. Your smartphone sends your current location to our server, which generates a waypoint file compatible with the drone's autopilot.
2. We upload the waypoint file to the drone and load your burrito in to our custom made Burrito Delivery Tube.
3. The drone flies to your location and releases the Burrito Delivery Tube. The burrito parachutes down to you, the drone flies itself home, and you enjoy your carne asada.

We built Burrito Bomber using a handful of open source projects and some new bits we created ourselves. All the code and 3D models we created for Burrito Bomber are on our GitHub page so you can build one too!

The airframe is a SkyWalker X8 Flying Wing. The plane uses Ardupilot to navigate the skies. The Burrito release mechanism is the combination of a Quantum RTR Bomb System, a 3" diameter mailing tube, and some 3D printed parts we designed in-house. The plane is controllable either manually via live video transmitted from the plane or autonomously using the Ardupilot autopilot. We use a Futaba 9C controller and EzUHF transmitter to manually control the plane. Check out the parts list for further details.

The web app is built in Flask. It gets the user's location via the HTML5 Geolocation API, generates a Mission Planner compatible waypoint file, then sends that to the drone operator. The drone operator uploads the waypoint file to the plane.

Unfortunately, Burrito Bomber as a commercial product is not yet allowed under current FAA guidelines. However, the FAA Modernization and Reform Act of 2012 requires the FAA to hammer out regulations for commercial use drones by September 2015. This means in 2015 we'll be able to take to the skies to bring you your burrito faster than you can say "¡Salsa roja por favor!".

(via Gizmodo. Thanks for the tip, Jason!)

We love the little Syma 107G helis, where are amazingly stable and cheap. They're just begging to be made autonomous, and now it's getting easier to do that with this hack. 

Hackaday reports:

[Jim] used a logic analyzer to do some in depth analysis of the Syma 107G helicopter’s IR protocol. We’ve seen work to reverse engineer this protocol in the past, but [Jim] has improved upon it.

Instead of reading the IR output of the controller, [Jim] connected a Saleae Logic directly to the controller’s circuitry. This allowed him to get more accurate timing, which helped him find out some new things about the protocol. He used this to create a detailed explanation of the protocol.

One of the major findings is that the controller used a 3 byte control packet, which contradicts past reverse engineering of the device. There’s also a new explanation of how multiple channels work. This allows multiple helicopters to be flown without the controllers interfering.

The write up is quite detailed, and explains the reverse engineering process. It also provides great information for anyone wanting to hack one of these low cost helicopters. From the details [Jim] worked out, it would be fairly easy to implement the protocol on your own hardware.

A Beautiful Sunny Day in Dubai !!

This H.A.L Hobby King quad was a joy to build. After fitting all the electronics we tested the CG and when all good then we tested the flying characteristics, we had to tweak the PIDs very little to get the Stablise going fine, Auto trim on CH7 works flawless and we managed to trim it to the finest possible. then by using the CH6 on my TX ( Futaba 14MZ) we couldnt acheive a good result for Loiter we are managing by the laptop ground station. this film was shot the day after we nailed the Loiter BUT its not perfect and we must try tweaking it further for the Loiter. Once we have the final figures I will publish it for H.A.L Hobby King. in general we are very happy with the quadcopter the price and the looks.

Hi, I previously set up a KKX525 with the APM2, but when one of my ESC's started fading, I thought I'd try something a little more upscale.

I have now put together a DJI 450F Flamewheel ARF including DJI motors, Opto ESCs and props from "RC-Drones" at a reasonable $180.00 and am quite happy with it at this point in it's evolution.

If you try this there are a few things you need to know, but no real problems.

Generally follow the downloadable instructions I have provided on the Arducopter Wiki for assembling and setting up and testing a quad copter and in addition see below.

1. If you get the ARF kit, it will come without instructions, but the manual can be downloaded from the DJI website under their FlameWheel support section. It is some help.

2. When you assemble the copter assemble the top board so that the narrow slots are in the front and the back and the wide slots are on each side.

3. Use blue or purple thread lock sparingly on all screws and try to keep it off the plastic. 

4. One of the nifty features of this quad is that the bottom board is also the power distribution board, heat the connection pads adequately to get a good wet bond with the ESC and Battery Wires.

5. You can cover the solder joints with liquid electrical tape or "Glyptol" if you can find it.

6. I tie wrapped each ESC under each arm with 2 tie-wraps front to back through the arm webbing.

7. You will need to add an external switching BEC, the DJI 30 amp Opto ESCs are great optically isolated ESC's but they do not include a built in BEC. I used a Castle 10 amp switcher for about $20.00 you can get cheaper ones from HobbyKing but you should use a switcher, they are much much more efficient. You do not need to cut or remove the ESCs power wire, it doesn't do anything.

8. You do not need to balance or tune the ESCs as shown on the Wiki (in fact you can't they are not programmable).

9. I recommend O-ring suspension mounting of the APM board between 4 standoffs secured at the ends of the long thin slots at the front and back of the top DJI board (assuming you put it in as I described in (2.) above.) Put a machine screw in the top of each standoff and suspend the APM board between them by putting a thin O-ring through each APM corner hole and passing the 2 loops over the top of the screw in each standoff. Size the O-rings so there is little or no APM board movement, they will still provide good vibration isolation. Too much free movement and they will cause handling lag. If you prefer you can mount a fiberglass or plastic board to the standoffs instead and Velcro or double side foam tape the APM board to it. I like the suspension method best though.

10. DJI supplies 2 sets of props, I recommend that for a 3 cell battery you use the longer 10" ones, Only DJI props will fit the self locking oval shaft of the motor unless you do what I did below. The shorter 8" props are for 4 cell use.

11. I didn't like the DJI props although they worked OK. They are light, flexible, noisy and would break easily. So I bored out some very nice GemFan 11" x 4.5 props that I was previously using on my KK to fit the outer diameter of the oval part of the motor shaft (about .31") and because of a thinner hub section I also had to add a washer under each prop. This is a very highly recommended modification, they are quieter, more efficient and will not break at the first unpleasant encounter. These are not the expensive carbon fiber Gemfans but the $2.50 carbon filled ones and they are still great, however looking at Innov8tive's web site I see they now have a 10" x 4.5 version of this prop already hubbed correctly for the DJI motors, buy these, they are a no brainer. My guess is if you use the stock ones you'll need new props soon anyway.

12. When starting the Flamewheel, it arms as per normal (throttle down and to the right) but if you do not spin the props within 10 seconds or so, the DJI Opto ESCs will shut down and you need to disarm (down and to the left) and then rearm again. Anytime you are on the ground longer than 10 seconds without turning the props you will need to rearm. This is not a big deal. You will quickly get used to it and given the capability of these things to shred flesh you will probably come to appreciate it as a worthwhile safety measure.

13. When you are setting up your board with only the USB connected, the ESCs beep continuously, ignore them.

14. When you turn your transmitter on and plug in the FlameWheel's battery, the ESCs should emit a short musical series and be quiet. You are then ready to arm and fly.

I have only covered substantial differences to the normal quadcopter setup procedure, but I highly recommend the Flamewheel 450, it flies quite well with standard PIDs but I will publish PID info later as I tweak it. (See Below).

Although the current stock PID's fly quite adequately, I have found the following changes decreased "twitchiness" and increased stability:

As of 12 / 8 / 2012

Reduce the Angular Rate PIDs P value from .175 stock to .145

Increase the Angular Rate PIDs I value from .010 stock to .030

Reduce the Stabilize PIDs P value from 4.5 stock to 4.0

The above will be revised as I continue to tweak.

Loiter will probably want some tweaking too and I will publish that after I have more extensive experience.

Roll and Pitch inputs are very sensitive so you need a light finger on the control sticks, but once you learn control with them I think you will be happy with them, If not you can probably detune them with the same PID's I am changing above.

I have a friend who has also successfully set up the Flamewheel F550 Hexacopter with larger motors and Graupner props following these procedures with such adaptations as are necessary with a Hex and he is using stock Hex PIDs.

Hope this helps, please feel free to ask questions I will respond as best I can.

Poker Flat Research Range (PFRR), managed by University of Alaska Fairbanks is working on several unmanned aircraft project.  One of them is converting surplus Raven UASs to run on the APM 2.5 and equip them with an extended battery pack. The Ravens will be used by researchers in Alaska for coastline, environmental and wildlife studies. 

As a proof of concept, I designed and built this carbon fiber nacelle for my undergrad research class. It would cover the APM should it be located on top of the fuselage to make more room for payloads. 

PFRR had a point cloud of the raven fuselage generated from a NextEngine 3D laser scanner.  From this model, I designed a mold that accurately met the surface of the fuselage.  After adding supports to the mold so it would not deform in the vacuum bag, I printed it on our Fortus 250mc 3D printer.  When the mold was finished, it was prepped with a mold release agent and 3 layers of carbon fabric were cut and infused with resin.  After curing in the vacuum bag, I removed, trimmed and sanded the part. 

On this first iteration, the part fit very closely with no large gaps.  Any gaps can be attributed to the rough trimming which will be simplified in later designs by adding a flange to the lip of the mold. 

All Suggestions Welcome!

The more suggestions that I get the more I will update the blog with the current progress of turning my F-117 Nighthawk R/C into a hybrid type glider.

Possible Ideas:

1. Cut off wings
2. Build custom glider wing or buy already made styrofoam glider wing
3. Once Complete with wings build a dome underneath plane that can house a camera
4. Add extra EDF Fans if needed for extra speed (its already extremely fast with only 64mm EDF fan house inside the plane in the middle

Friday morning

Friday evening...

I finally managed to get access to a decent camera on a day with good weather! The following video is mostly demonstrating the quadrotors stability under the new four-state Kalman filter I have implemented, that estimates both the current pitch/roll and two gyroscope biases. Most of the flight is hands-off, with the occasional input to correct for wind and inevitable position drift.

    I feel this angle control is pretty much perfect, so I will now be moving on to integrating extra sensors to allow for direct velocity and position control. I’m excited to see just how complex I can make the navigation system before the Pi cracks, but it seems for now despite using unoptimized  double precision floating point math for everything it still has plenty of headroom to go!

    The (rather messy) code is available at https://github.com/big5824/Picopter. Now that the first stage is finished I can start going through cleaning up the spaghetti of test code, but that will have to wait for now.

http://www.adafruit.com/blog/2012/12/07/a-quadrocopter-powered-by-raspberry-pi-piday-raspberrypi-raspberry_pi/