This will occur when you have mixed cations (here Nd and Dy) or mixed anions on the same cristallographic site. Their occupancy need to be refined (so occDy + occNd = 1) for the concerned site, before creating the .cif file. The size of each part of the pie chart will be directly proportional to these refined occupancies, and will be automatically displayed on Vesta. See the case of BaBi4Ti4O15 attached.
You just need to refine the occupancy when refining the structure (add some constraints occ.Dy + occ.Nd = 1) on your software (Fullprof, Jana, etc)
PS : If the .cif file you download does not have refined occupancies you will not be able to display these pie chart.
Generally this representation means that you have a statistical (non ordered) distribution of Dy and Nd over the equivalent positions, along the cell (some are 100% Dy, some 100% Nd, but both cations are not on the same sites at the same moment), leading for example for your structure :
On Site 1 (light brown), you have 6 equivalent positions, meaning about 5 of them are Nd, and 1 Dy.
On Site 2 (dark brown), you have 6 equivalent positions, meaning about 4 of them are Nd, and 2 Dy.
(I can not see properly the sites on your picture but I do believe it is ok).
However it is not possible to have "more information" (you can not know their exact distribution pattern, you can not know how the Cation-Anions distances are evolving as everything is averaged, etc)