I am not sure how well the tech will hold up to the effects of wind, car engine sounds, screaming kids, customizations (hold the pickles), accents, etc.
What I think will happen is that they will roll out this technology, and then be faced with a bunch of irritated customers re-doing their order at the payment window because the voice AI completely screwed up their order.
I think mobile ordering through an app is a much more viable option.