I'm a little unclear as to what you're actually trying to do. But I'm going to go ahead and say that it is almost certainly possible to achieve in FME.
There are a number of transformers available in FME that could help you with your task. Firstly, there is the NeighborFinder. You can use this to find features which are close to each other.
Finds the Candidate features closest to each Base feature and merges
their attributes onto the Base feature according to the Accumulation
Mode parameter. Or, if Input is set to Candidates Only, performs the
same operation using each Candidate as a Base feature, considering all
other Candidates, but not itself.
There is also the Snapper transformer and the AnchoredSnapper transformer to snap the points to a particular position.
From the Snapper Documentation:
Brings lines, segments, end points or vertex points of features
together if they are within a certain distance of each other and
(optionally) if they have one or more attributes in common.
From the AnchoredSnapper Documentation:
Takes a series of features that match the input specification and
performs snapping on the features that lie within the specified
tolerance from other features that match the input specification. You
can use this transformer to perform cleaning operations on data during
Also, you will probably want to use the FeatureMerger.
From the FeatureMerger Documentation:
Copies and merges the attributes/geometry from one feature (or
multiple features) onto another feature (or multiple features).
There are many other transformers that may help with your task, depending on how you want to tackle the problem. But these should steer you on the right track. There are also pages which discuss clustering methods in FME. Such as here, here, and here.
Hope this helps...