A note from Bill Bennett of Pizzeys (Australia seems to be adopting the standard from In re O’Farrell just as the US courts are distancing themselves from it):
We have previously flagged that the APO might modify their practice in relation to the “obvious to try” standard.
We have now observed that the examiner’s manual was amended on August 2nd. (This can be found by clicking here.)
According to the APO, an invention will be obvious to try if there is a “reasonable expectation that the solution might well solve the problem”.
This appears to echo the High Court’s decision in Aktiebolaget Hässle v Alphapharm Pty Limited  HCA 59, which expressly endorsed the reasoning of the US Court of Appeals for the Federal Circuit in In re O’Farrell (853 F.2d 894) as the better approach under Australian law.
The Australian Examiner’s Manual now recognises that Australian law includes the requirement that anything suggested as “obvious to try” must carry with it a “reasonable expectation of success” in order to make the invention obvious. As in US law under O’Farrell, rebuttal of the factual assertion that any anticipated success is “reasonable” may therefore help overcome objections that the invention is obvious because it was “obvious to try”.
In practice, Australian examiners may not be so bold as to assert “obvious to try” as the sole reason for a claim rejection in view of the overarching principle that the skilled person must be “directly led” to the invention for it to be obvious.
Indeed, recent examination reports are asserting that references “provide motivation to make the invention” and “directly lead the skilled person to the invention” and only then stating as a consequence of those direct leadings that a “reasonable expectation of success” has been established.