What is the state estimation used for in the DWL paper?

[alstondu]

Hello, it seems that the estimated state from the DWL is not used by the actor, hence not participated in the real robot deployment. I wonder what is the advantage of designing this estimation process?

[hxydebug]

Maybe the DWL is not included in this github code