The roadmap

The UAV project has several milestones on the road to full automated capabilities. The stages that have been identified so far:

• Build the prototype

  The LM-110 was identified as the prototype airframe. It took roughly thirty hours to construct, including a long nap and waiting for the fuel proof dope to dry.

• Learn to fly under manual control

  The project's chief test pilot is still learning to hover the prototype with a standard RC helicopter control. Roughly two liters of nitro fuel have been run through the engine and the aircraft is performing quite well. A crash on 20 September 2001 broke the tailboom and tailrotor gear box, as well as lost some of the tail rotor linkage. Replacements were ordered. The tailboom and gearbox was repaired and some of the linkages refabricated. It isn't pretty, but it flies again.

• Manual control via computer

  The first step towards computer control will be manual control via a laptop, the SSC servo controller and visual observation of the prototype. This might be easier than the the previous step since the computer control can do control mixing and "hover throttle" settings that the manual control does not. The helicopter was tethered to the laptop rather than risking the embedded system at this point.

• Manual control via computer with video feed

  We are here. The xcam video unit has been fitted and the aircraft balanced for the extra weight of the camera and sender unit. A public demo was given in early October 2001. The prototype was capable of taking off and hovering, despite the extra 100g of payload.

• Manual control via computer and EFIS

  Once a low-cost AHRS system has been found and installed, the next step is to learn to fly the prototype solely on the instruments. Similar to flight in IMC, the helicopter's attitude will be controlled by the test pilot who only views the EFIS screen. A spotter will watch the helicopter to ensure that it does not cause any sort of damage to persons or property. Again, the helicopter will likely be tethered. A screen shot for the EFIS code is shown on the left.

• Automatic control via computer

  This is the next to last step towards the ultimate goal of the project. Once the prototype is controllable via the EFIS, the control software can start incrementally adding support. Tail rotor control is likely to be the first extra functionality in a fashion similar to the "heading hold" gyros. Then an automatic leveling function for the main rotor. Add a sonar altimeter and be able to land or autorotate automatically. And lastly a "position hold" feature that will combine all of the stabilized functions to allow the UAV to stay in one location.

• Implement mid scale aircraft

  The stage two prototype is likely to be a commercially available airframe of sufficient size to carry the onboard systems (computer, AHRS, GPS, servo drivers, sensors, etc), but well short of the target payload. The Observer (pictured on the right) from Bergen is a good candidate. However, a Kyosho Concept 60 became available for the right price, so it has been purchased. See the prototype document for more details.

• Handheld control

  The command and control software should be fairly robust at this point, so the joystick backup can be discarded. The ground station can be collapsed to run entirely on a handheld computer, such as a Compaq iPAQ with Linux and an 802.11 card.

• Design and build large aircraft

  Once the command and control software has been debugged on the prototype airframe, the full size rotorcrft needs to be designed and built. Much of the hardware from the prototype can be reused -- the servo controller, the laptop ground station, and the AHRS. The design of the larger airframe is an open problem and one that will require input from potential users. More ideas on the overall design