From a 2011 paper G.Keady's Erdos number might become 2:
Erdos - 1. Vijayan, K. - .2 Keady,G. (2011 paper)
though this is an "applied maths" paper and might not be listed in Math Reviews
With more mathematical (Graph Theory) papers one way to 3 will (after 2016?) be
Erdos - 1. Caccetta, L. - 2. Achuthan - 3. Keady,G.
where Achuthan is both of them, Nirmala and "Arch", N.R.
(Preprints of the papers with Achuthan are up on arXiv.)
Prior to 2010 G.Keady's Erdos number was (at most) 4:
Erdos - 1. Reznick,B,A. - 2. Fitzpatrick,S. - 3. Calvert,B. - 4. Keady,G.
Erdos - 1. Reznick,B,A. - 2. Fitzpatrick,S. - 3. Dragomir,S. - 4. Keady,G.
Erdos - 1. Caccetta, L - 2. Wang, S - 3. Dragomir,S. - 4. Keady,G.
(I have a Maple review with Simon Fitzpatrick, but doesn't really count.
Rules for calculating Erdos numbers are papers listed in Maths Reviews.)
*SOME* people at UWA and Curtin Maths with small Erdos numbers smaller than mine:
Caccetta 1
Vijayan 1
Achuthan, Nirmala 2
Achuthan, N.R. 2
Fitzpatrick 2
Praeger 2
Wang 2
Erdos Number 1 individuals are at
https://files.oakland.edu/users/grossman/enp/Erdos1.html
and one can find many co-authors there.