Semantic machines and traffic systems

I propose the semantic traffic cop, literally a web bot that checks out autobots around town. So, some bot wants to come to town, the cops send in a web bot to make sure he obeys.

Build a traffic layer using B-Tree/BSON standard, a literal system to manage traffic on closed environments like, say, the ultra speed bus rapid transit project. All events seen from the point of view of any real bot is standard BSON Tree nested form, ready for convolution with the local police graph. The result is the inner join of their mutual rule set, the set of possible nested orders after the next move, a predictive semantic graph of things the two binary parties agree.