Ria was an interesting project that taught me a great deal. It got me back up to speed in actually assembling Lego® models. It allowed me to design a model that ran accurately enough to do complex ball sorting routines. To an extent, this brought to life a simulation that has sat on my laptop for several years, which was itself satisfying.

But the most valuable lesson was that my biggest satisfaction came before I added the grab and spent time working though the complex software routines. A few parts, one motor, one sensor and a very few lines of code. Voila, a complete train system. I sat and watched it. Then I watched it some more. It was the best phase of the project. Time to move on.

Meet Sugar. I hit the nail on the head when I said that Ria could implement a Turing machine. I spent ages trying to solve the problem of using the light sensor for sorting and for motion control, but in a beamrider it's not so easy. The structure that does make it easy, though, was staring me in the face. Sugar will revisit the QuAD problem, with new eyes. Suspend the unit, put the grab underneath and point the light sensor downwards.

Compared with real monorails, aerial trams or telepheriques, the are problems with both my and Anthony Fudd's approach to the problem. Most true telepheriques and, I gather, aerial trams in use in the US, aren't actually self-powered: instead they act like cable cars, fixed to the cable. With long enough flying leads, this would be a great Lego® solution as well, but I don't have long leads yet. Self-powered monorails can usually be lifted off the track or cable, for maintenance if nothing else. And they can pass over suspension poles as well. Neither my tram nor Anthony's will do either thing. Additionally, mine isn't accurate enough for complex work, and Anthony's won't run on a tight cable, which limits it's potential length.

So, the next project is to design a unit that will (a) run accurately on a taut cable, (b) pass over suspension poles and (c) mount and dismount easily on the cable, without needing to take advanced crochet classes. It turns out that most solutions to (b) and (c) are the same. So, Sugar will:

• run accurately on a taut cable;
• pass over suspension poles, mount and dismount easily on the cable;
• suspend the RCX beneath the cable for stability;
• point the light sensor downwards for control;
• mount the grab underneath;

Why Sugar? Well, I wrongly remembered the plantation monorail as being for sugar cane rather than bananas. Since the projects have started "Q" and R, the next should be S. I'd done the .gif (just as the first stage in a big military project is to design the badge) before I realised my mistake. Anyway, that aside, let's go:

The mechanism I'm going to use suspends the RCX and grab from two wheels that run above the cable. The wheels will be cantilevered from one side to hang exactly over the centre of gravity. Everything else will hang below the cable. Actually, for the first stage, I have a piece of old curtain rail which will attach to one face of Ria's track to provide a rigid test track, so I'll start there. I have future ideas for a more complex Sugar which I'll try to keep in mind during development, though.