There's some very dull footage (link below) and my description will be way too technical to interest most of you. In short, this email could generate an enormous yawn. Suffice to say that controlling the launch of my preferred kite from KitetiK, with the long in gestation winches/drills/RC with no crew and with it being very easy, is a major milestone in developing KitetiK.
Those of you already yawning should neither read further nor watch the video (just over a minute). If there is anyone not in that category, read and watch on...
The SS (single skin kite) is rigged with a 4:2:1 mixer - this is a few lines and a couple of blocks. The details are not important. Suffice to say that the mixer helps keep the shape of the foil as the trailing edge is sheeted in. I've performed surgery on the mixer so that the middle of the span, about 6sqm perhaps, is not attached to the mixer - so this part of the kite is effectively sheeted in, though not the brake on the trailing edge itself, on just the front lines. The lines leading to these bridle lines are connected directly to the front flying lines. These are winched through separated tow point blocks so that there is less than 3m of bridle between tow point block and kite. With the tow points about 2.5m apart and across the wind, when the kite is launched onto these short lines, it flies very stably - albeit with the wing tips flogging.
Two spooling winches are driven by 18v brushless drills themselves controlled by a standard hobby RC transmitter through some Arduino logic. One controller lever, the reeling lever, either drives them both in or both out. Another steers (ie drives one line one direction and the other the opposite direction). An RC toggle switch converts the steering from one mode to another. In one mode, the steering signal is interpreted as a position relative to neutral, where "distance" is measured by a rotary encoder on the chuck of the drill. This allows movement of about 1m (opposite directions on each drill, so effectively 2m line differential). In the other mode, the steering signal is interpreted as a speed to drive the drills in opposite directions (positive and negative for different directions). This is necessary as, with differential load and winding onto the two winches, this second mode is used to get the lines "even". The lines have paint on them every 6m or so to allow evening up.
When flying on the very short lines, the winches are used to ease it out to full bridle length - that's about 6m. At this point, the rest of the front side of the mixer is connected to the flying lines (ie the ones on the winch and carrying the centre of the span). The rear lines are still slack. While they are slack the wing tips flog. (It is necessary to have these two connections to the flying lines separated as the mixer would not fit through the blocks.)
With the tow points for the two front lines separated across the wind, there is a natural steering input towards zenith - if those of you who are kiters who remember some high school geometry draw some pictures, you'll work it out. Unfortunately, however, without a damping term, as the line length increases this steering input generates oscillations that slowly build - without active control this will end badly. At short line lengths, the oscillations are minor and do not build. Hence it is dynamically stable at short line lengths but unstable, albeit taking a while for things to go bad, at longer line lengths. (Apologies for use of terms like "fairly stable" in the video - it's not stable, but goes bad slowly!).
With all front lines/bridles connected, the kite is eased out to flying height. This was one very satisfactory aspect of this experiment. I would ease the kite perhaps 2m or 3m at a time. If oscillations started, they could be very easily damped to near zero with minimal steering input (mode 1 steering). Mode 2 steering was employed a couple of times to rebalance line lengths. At 25m (perhaps I should have gone higher...but there was another boat moored downwind), I kept it at zenith very easily for a few minutes. It does require active piloting - but not very much. I suspect that might be possible to automate.
If I had crew and was going sailing, the tow points would be moved to the sailing tow point position in the middle of the leeward (long) hull. This has been demonstrated previously and is relatively straight forward. The lines can be cleated there. In at least the short term rear lines will be eased off a bar by a crew member while the kite is being reeled out to flying height. When the kite is at sailing height, these rear lines will be ready to use manually. Further down the track the winch lines may be transferred from flying lines to rear lines - though a mechanism to do that, while keeping the kite at zenith, has not yet been developed.
As the kite is a prototype with knotted rather than spliced bridles, there is a higher likelihood that the bridles in the wing tips will be tangled while flogging at this stage. Although splices would improve this, some mechanism to stop this generally needs to be developed. I have trialed a sock over mixers before with some success - but that is quite a different problem.
Retrieval is essentially reverse of launching and reeling out. That was all achieved quite easily. A reefing line runs through the ribs in the centre of the span. So the whole kite is easily doused when back flying at <3m - I didn't capture that on the video.