The xbee is remoted so that it can be mounted below the plate, but the rest of the electronics stay above the plate (in particular the GPS needs to see the sky). On the left is a FunnelIO board with a perfboard mounted, in the middle is the same IMU as the QC, but without accelerometers or gyros, attached to it is the GPS and altimeter. (GPS not actually attached for clarity).

The FunnelIO board is going to be replaced with a smaller custom board once the system is rung out. Details to follow.

And the Quad Copter is in progress– really just doing software now, but on hold while I finish up the powered parachute (which I need first):

Software’s a long way from written, but this is the first look at the controller for the helicopter — The red thing is the 6DOF — it’s three axes of gyro sensors and three axes of accelerometers to determine the precise orientation and movement of the helicopter in flight so the main computer can figure out how to orient itself. Very slick product from Sparkfun– what I particularly like about this one is that there’s an on board atmega168– so all of the Kalman filtering and preprocessing is going on in the sub-board, and then the processed data is uploaded to the main board.

The rubber band (or gumband), I must say, is only temporary while I put together the elastic system to support the battery. Everything else is hot-glued down to the 3/32″ hobby plywood. The whole thing is going to be attached on top of the central core of the helicopter with screws and possibly some rubber or cork to dampen vibration from the four motors.

Both sets of controllers are based on the FIO platform — it’s an arduino-like device, notably scores a built-in LiPo charger and Xbee Socket. Since I’m using Xbees for my radio communications, it makes it an easy choice. For the helicopter, which has a lot of sensors to integrate, I’ve had custom boards made through BatchPCB. These solder up to the back of the FIO board, and provide plug-in locations for all of the sensors and connections. The powered parachute will have perfboards attached to its FIO boards and wired up that way, rather than custom boards, since it has far fewer connections to make.

For the helicopter, the landside controller has two joysticks, an OLED display, and a 5-position “clicky” joystick for menu navigation. The airside controller has the following sensors:

• Atomic 6DOF
• Digital compass
• Digital pressure sensor
• GPS
• Ammeter on flight battery
• Simple voltage sensor on flight battery

The helicopter uses the 2.4GHz Xbee-Pro for communications, with a typical range of 1 mile.

The powered parachute has a somewhat simpler controller — a slide pot for the throttle, which controls altitude as well, a dial for steering (which controls both the parachute’s steering lines and the nose gear), a pushbutton for the camera shutter, an LCD and a hotpot for menu navigation and control. The camera, since it points downward, will mostly be in an intervalometer setting, but the period between the photos will be determined by the altitude and speed of the aircraft.

Airside on the powered parachute is a similar suite of sensors, although missing most of the “big ones”:

• Custom built pressure sensor (16bit)
• GPS
• Ammeter on flight battery
• Simple voltage sensor on flight battery

And that’s all for now — photos and details forthcoming as devices are completed.

Dorothy 6 is a helicopter with 4 electric powered rotors– sensors on board keep it level. It’s radio controlled, using custom electronics and Xbee Radios. D6 uses the 2.4GHz Xbee-Pro radios for a maximum range of about 1 mile. As the battery power only lasts for 15-20 minutes, 1 mile range is plenty. The helicopter format is ideal for getting in close to objects, and shooting panoramas as the camera platform can remain stationary locked onto GPS coordinates. If there’s a loss of power, D6 will trigger a parachute to land relatively gently. You can see some videos of the parachute tests on Vimeo, here and here. (The second one is only available to my backers on Kickstarter)

The Flying Machine takes its design from ultralight aircraft. There’s a parachute-like canopy over the base, which has a single large electric motor pushing everything forward. Because of the large wing area, it’s a very slow, but very efficient craft. Flight times are expected to be 60-90 minutes. However, it’s always flying forwards. This makes it much better suited to large area surveys. The initial camera configuration will have the camera just pointed down for orthographic shots. Later additions will add a pan/tilt rig to it, but at 10-15mph, it’s still going to be suited to taking long panoramas of things. Luckily, most riverside industrial sites are long and skinny. The radios for these are the 900MHz Xbee series, with a 6 mile range. Because it’s already hanging from a parachute– incase of a power failure, it will glide to a landing, rather than dropping from the sky.

Both UAVs will be capable of taking film or digital cameras, currently either a Canon A590IS running CHDK or a Nikon N70 with a 28mm prime lens, shooting generally onto ISO100 35mm Fuji Reala. I have plans to upgrade equipment, likely digitally to a micro-4/3rds camera, and to a motor drive medium format film camera– probably the Pentax 645 system. (You can help facilitate that by supporting the project on Kickstarter, of course).

Build for these is just beginning, so expect to see posts with photos in coming weeks.

Yellow = unscouted (all at present)
Blue = scouted and good
Red = scouted and no good (e.g., nothing at site)
Purple = preliminary imagery available.

My ultimate goal is to foster a new understanding of our industrial heritage through aerial photography, providing new perspectives on what we see every day.

As tools for aerial photography, helicopters and airplanes suffer from similar problems– it is rarely safe to fly them below 500 feet, especially when flying near buildings and other structures. Depending on a project’s size and budget, various solutions have been utilized to acquire aerial views: Photographers from the Historic American Engineering Record, while documenting the Homestead Works, climbed the water tower to get a view from above; cherry pickers and tall ladders work for many jobs; on other jobs, carbon-fiber poles up to 40′ long can be useful. To get a higher perspective, between about 50 feet and 500 feet, requires a very different kind of technology.

For two years I’ve been working with kite aerial photography. Invented by a Frenchman in 1885, this method has been practiced on an off for over a century, but has only really taken off with the advent of inexpensive digital cameras. The advantage of KAP lies in the ability to get very close to a structure without risking damage to the camera or the structure itself. The only equipment required is a kite and a camera, suspended a few hundred feet below the kite.

We estimate that there are about a hundred serious KAP photographers worldwide, including roughly a dozen in the San Francisco Bay Area, where my interest in KAP began. The Bay area has the perfect combination of smooth winds and open spaces to perfect the craft; Pittsburgh, however, has neither of these, particularly near the sites I am interested in studying. Between hills and power lines, a different technology is required to work here.

I am currently developing an unmanned aerial vehicle (UAV) for work in Pittsburgh. This will be a small helicopter-like device, about 2 feet square with four battery powered rotors. It will have a slew of sensors that permit level flight without input, allowing the operator to simply tell the UAV where to go, rather than controlling every aspect of its flight en route. This will permit photography of abandoned sites without risking harm to myself, the camera or the structures being studied.

In California, I focused my KAP efforts on documenting the Alameda Naval Air Station, which closed in 1993. I also studied various historic buildings in the Bay Area, including a condemned World War II office building. The upper floor of this building has collapsted in a fire, and no maintenance had been done to the wooden structure in decades. The three-story structure was covered in graffiti and slated to be torn down; without the ability to photograph it form above, it would not have been documented architecturally before it was burned to the ground by vandals in early 2009. These photos are collected here.

Here in Pittsburgh, I have approximately 40 sites on my short list to study. This list was compiled largely with the help of the Rivers of Steel National Heritage Area.

The project is broken into three phases:

Phase One Continued research into these historic industrial sites, combining it with aerial surveys to assess the status and historic value of these sites. The end of this phase will whittle down the list of sites to 6 of primary interest.

Phase Two I will conduct detailed surveys, including high-resolution orthoimagery of the 6 sites, as well as close-up images of industrial details not visible from the ground.

Phase Three I will compile the photos and make them available to the public in several ways. This will include mapping from the imagery collected — comparing them to the historic plat maps — as well as producing a photography show, a book, and a website with the images and data freely available (licensed under Creative Commons)

The project is currently planned to begin March 2010, with completion around March of 2012. The current process is researching existing sites around the area — see map here — and looking for funding sources.