[spidey]

I do exactly this as part of heeling. I don't see why you need a special cue for it, that seems to me to be making things more complicated than they need to be - telling the dog to "heel" should mean "maintain/assume heel position, sit when we halt", adding another cue might just serve to dilute your original cue (I also think that maybe you haven't truly established what heel really means in the dog's mind if you have to come up with a separate cue to get the dog back into position - it seems to me that if the dog genuinely understands the cue, then you should simply have to issue the cue before you move off and the dog should assume and maintain position, no matter what direction you move in). I make a point of having my dog heel through about-turns in both directions, spins on the spot, and side-to-side (mainly because I'm building up to Rally competition, and there are Rally exercises which involve this sort of thing, and also because heeling is boring and stressful if you don't make it interesting). I also wonder if you always reward/release at the halt/sit, rather than during the heeling itself - in my opinion, if you always treat the halt/sit as the "end" of the exercise, then the dog first of all learns to "switch off" a bit when you halt, and also learns to view the goal of heeling as the halt/sit, rather than maintaining position. For this reason I'm working backing up, about-turns and sideways movements into my doodling patterns, and I almost NEVER reward/release at the halt now that we've established the automatic sit, I reward far more frequently for maintaining correct position while moving.