The majority of modeling efforts in problems of nanotechnology are based on numerical simulations. The common reasonale view prevails that the behavior of nanovolumes cannot be described by usual continuum physics models which have been tested and traditionally applied at the macroscale. Generalized continuum physics models, on the other hand, contain too many phenomenological coefficients, difficult to interpret physically and determine experimentally. In recent years a "compromising" approach has been advanced by the author and co-workers which seems to be quite promising for nano-engineering applications. It is based on enchancing the standard continuum mechanics models with deterministic gradients, as well as stochastic terms, which allow for an effective modeling of nanoscale phenomena and experimental observations which cannot be captured by deterministic models alone or by corresponding atomistic simulations. This is shown by example problems fordeformation and mass/heat transport processes at the nanoscale. Corresponding models of nanoelasticity, nanoplasticity and nanodiffunions are developed and discussed.