I'm not sure which era you're talking about exactly. Graphene and carbon nanotubes can't currently be made both big and perfect, and are lame when imperfect. Nanoscopic robots have problems with sticking together and jumping around due to brownian forces, and also are just very hard to build. Chemical-based robotics has been a crapshoot because quantum chemistry is hard. The last one has been tackled with machine learning pretty well recently, where natural biological analogues exist.

As a result, about as far as we've gotten is nanoscopically fine dust. It has uses, but it's only a technology the same way pea gravel is. It's looking like a lot of the stuff nanobots were supposed to do is going to fall to biotech instead.