Our main goal will be to demonstrate that we can control the component separation to a degree that allows us to obtain competitive constraints on the tensor-to-scalar ratio r.
From my point of view, there are two very interesting challenges: First, the extragalactic component separation will need to be separated into the CIB and the CMB signal. Since the CIB is strongly mixed with the signal from the Galactic dust, this will become a very challenging ‘three-body problem’.
Second, in order to model the dust and synchrotron component from our Galaxy in the most accurate way possible, we need to account for line-of-sight variations of the components such as frequency decorrelation. Even though no evidence for this effect can be found in Planck data, we will need to deal with it in more precise experiments. Here, the HI data (HI4PI, GALFA) and and 3D dust data will come in very handy, probing the radial velocity and distance axis of the ISM, respectively.