…or it should rely on continuous reception of a carrier, or a burst of data every xx milliseconds, and stop (but then SURELY stop) if these aren’t present.
This is a nontrivial problem because something as simple as the lever breaking off a switch would result in a disaster. However any powered tug has that same problem…
I would definitely implement some crypto on the data (to take care of interference) plus a separate carrier / loss of carrier detection.
One would definitely need some product liability insurance 
and that’s not a theoretical risk…
I’d have thought that a regular 2.4ghz radio modeling remote control would have all the necessary protocols in place – after all some of the big jets and helicopters can and do cause fatalities when things go wrong. They’ve become very reliable and glitches are almost unheard of. They already have failsafe mechanisms designed in. I believe Futaba’s major business is actually commercial remote control, with the radio modeling as an offshoot. Some of the chips used are commercially available:
e.g. google cywusb6935_8.pdf (used in Spektrum systems)
a burst of data every xx milliseconds, and stop (but then SURELY stop) if these aren’t present.
Lazy Sunday evening muse: if I could get myself to concentrate on this, and think really hard, I might well be reinventing the watchdog timer.
One like these: https://www.acairtechnology.com/Products.aspx?ModelID=3
Looks very cool. Almost portable at 20-odd kgs. But 4k USD?
I am sure an electronics whizz (hi Peter!) could build one for less.
Thoughts?
I suspect one doesn’t even need to know any electronics – a drive unit from an e-bike will probably do. You just need to increase the thrust and lower the speed; for a hub motor, it only means using much smaller wheels (e.g. fitting a tyre directly on the hub).
I am sure an electronics whizz (hi Peter!) could build one for less.
Sure one could. It would be an interesting project.
I would definitely want an INHIBIT switch working over a second totally separate radio channel though Anybody even vaguely familiar with long range radio controlled models (e.g. FPV, especially the, ahem, not so legal FPV like FL200) will know that approximately 99% of them get smashed up eventually, due to loss of signal / interference over the link.
The interesting point is that IF you can get the thing under the nosewheel, and thus have the benefit of the weight pushing it down, a tug weighing almost nothing would still work fine. 20kg is not portable IMHO.
You just need to increase the thrust and lower the speed; for a hub motor, it only means using much smaller wheels (e.g. fitting a tyre directly on the hub).
You need bidirectional variable speed drive on each track, but that’s just two servos. Plus something to sort out the bit that grabs and lifts the nosewheel – that bit won’t be trivial. It’s normally done with a belt which is manually put around the nosewheel oleo and then (using a motor driven winch) pulled up, and that lifts the nosewheel onto the platform.
Well I guess you do need electronics to know how to set up the motor controllers and link them to the remote receiver (i supposes you can’t run that sort of amperage through the receiver directly ^^) otherwise it only goes in a straight line…
I think the nosewheel IS supposed to get into the cradle thingy anyway. I wonder what the weight is. Batteries?!
I would definitely want an INHIBIT switch working over a second totally separate radio channel though
We know you like redundancy (except with slip indicators ), and it would certainly not be a bad idea, but why would that be a requirement? You can just drop dead if the radio link goes haywire, just make sure it is encrypted and nobody can do a MITM
Prob99 you will be using an off the shelf RC transmitter (maybe repackaged to justify the overall price tag) and if one of the pots in that goes O/C at one end (pretty likely – they are crap quality) then the tug will go flat out.
I guess they have considered this and put in some sanity checks / slew rate limits etc.
I do have four slip indicators BTW
