FairIP index

Ground-based AWE Automation with Human Supervision

A prominent AWE school of thought is focused on flying-robots so wonderful they operate reliably for long periods unattended. It might be the first aviation technology to do so, but in the field of Artificial Intelligence its been clear for a decade at least that truly smart systems are far slower to emerge than was expected. Its 2010 & there is no HAL9000 or Terminator in sight. A new appreciation exists for naturally evolved intelligence.

Robots still have vast promise for AWE. They do work properly in fabulously expensive programs like space exploration & highly structured environments like factories. They work well under close supervision with humans to relieve tedious chores. Even the simplest autopilot is a pilot's friend, much as "lashing-the-tiller" allows a sailor relief, but no autopilot ever developed comes close to the senses, knowledge, & wisdom of an expert human pilot.

Thus practical AWECS presently requires human supervision. The workload on the human must be kept light, so the system should be self-flying with the human ready for tricky operations & exception-handling. All auxiliary automation is best kept on the ground where it is not mass-limited or radio-linked. Ground equipment can be serviced easier, is far cheaper, & less exposed to loss than flying equipment.

Basic ground-based AWE mechanical agents work far better scary flaky flying-robots. The ground-based agents can be redundant combinations of well proven industrial control technology too fragile or massive to fly. Some ground-based agents will be semi-autonomous tow/winch vehicles on tracks or following buried wires. Smart-winches will handle lines like a pro, actuating the array by merely "pulling the strings."

Ground-based radar is a clear winner for position tracking of aerial arrays.  A single rotating "eye" can see every flying object provided with a simple cheap corner reflector in all lighting and weather conditions. System backups are cheap-and-easy if ground-based. Hardly any electronics need fly for AWE to work well and scaled up. Even the navigation lighting function might be ground-based, with on-demand floodlights brilliantly illuminating an array's membranes and reflectors,  if air-traffic strays.

Ground-based automation with human supervision is the safest AWE bet Its a more insurable investment. The "kite-pilot" jobs are very cool. There will be less loss of life & the technology will be more accessible at lower cost.

CoopIP/FairIP                           ~~Dave Santos                          May 15, 2010                  M1526

Comment and development of this topic will be occurring here.

Terms and aspects:    Airborne Wind Energy (AWE), artificial intelligence (AI), robots, robotics, self-flying, tricky operations, exception handling, auxiliary automation, automation, human supervision,  mass-limited, mass-unlimited, radio linked, aloft equipment, ground-based equipment, exposure to loss of equipment, ground-based AWE mechanical agents, equipment service, flying robots, industrial control technology, ground-based radar, position tracking, aerial arrays, reflector, system backups, on-demand floodlights, ground-based automation, avionics,


Fine analysis.

Possible studying steering systems according to scale and use of AWECS:
-complete automation
-no automation
-active control: automatic control or human control (for example for cell phone charger as toy) 
-passive control (KiteLab' searches).

An example of mix of complete and semi-automation and ground-based and flying based (pod) automation:
Pod advantages are allowing only one line, and also an easier steering because control lines are close to the wing. That compensates inconvenience of aloft avionics.


"To escape fantasy-land, major invention starts as a working toy."         ~~Dave Santos
SkySails has done many things right-
Choice of Supervised Automation/Human Piloting
Adoption of mature parafoil tech rather than betting on kiteplanes, kytoons, etc.
Logging high real-world flight-hours to identify & eliminate failure-modes
Compare with similarly well-funded AWE efforts stuck year after year in "promise mode."  SkySails is still dependent on radio-links and avionics aloft, which are far less essential to multiple anchor-point kite-fields with multi-line control.

Dave Santos                              May17, 2010

I will try to develop a kite control system at TU Delft, beginning the
1st of July. I might need four years, but I am optimistic, that we can
publish our results in an open source manner.

I have some safety-critical knowledge in the field of the development of
real-time control systems, because I was working in the field of medical
device development for five years.

Best regards,                      ~~Uwe Fechner                     May17, 2010