Direct imaging surveys, like the one carried out with SPHERE, are targeting young planetary systems, and are relevant to establish the initial conditions of planet formation. The youngest systems are in fact the most difficult because planets are still embedded in gas and dust even though we expect them to be very bright. 

In that context, AB Aur occupies a special place as being young, and with a long record of detailed observations at various wavelengths. At sub-millimeter wavelengths, ALMA has detected two molecular spiral arms at less than 100 au, together with a suspicion of 1 or 2 protoplanet candidates proposed at the tip of spiral arms (Tang et al. 2017), echoed by a recently claimed protoplanet candidate at this very same location (Currie et al. 2022).

Since Decembre 2019 we have been monitoring AB Aur with SPHERE at the VLT in the near IR. Polarimetric and non polarimetric images identified a wealth of structures in scattered light at many scales, in particular two spiral arms clearly overlapping with the ones from ALMA. The gain in angular resolution has allowed us to discover a bright feature at 0.18'' (about 30 au) from the star, reminiscent of the connexion between the inner and outer spiral arms as expected by the theory of wave density, when such waves are launched by planets (Boccaletti et al. 2020). This is a plausible evidence of a protoplanet signature. 

With now 3 epochs of SPHERE high quality polarimetric observations spanning ~4 years from 2019 to 2023, we can definitely confirm the motion of this "twist" structure by several degrees. The whole disk is rotating, very clearly, at an angular speed matching the local Keplerian velocity, hence possibly at odd with respect to the standard representation of density waves. Several shadow features are also evolving at different rates revealing the impact of the inner disk on the outer disk illumination. We will also present Halpha imaging obtained with SPHERE/ZIMPOL aiming at identifying accretion signatures.