When studying the position, we found that the underlying pawn ending is an easy win for white. The underlying minor piece ending is a win for white too, allthough not so easy. Both endings follow the same scheme. The scheme is to attack on both wings. Black wants to defend on both wings, but he can't keep up since his pieces are bound to control white's outside passer. So effectively you can outnumber black on the kingside, nibbling away the pawns and creating a second passer there. That will proof to be to much.
You have to be aware of blacks pawn sac c5, which can be played at an inconvenient moment. That will tear the a and b pawn apart, so they become more vulnerable.
On the kingside you must be prepared to induce weaknesses. In some cases you will need to sac a pawn yourself for that. In the case of a minor piece ending, your key pawns will be perfectly safe from the bishop on the dark squares.
The same pawnending with the rooks added in stead of the minor pieces proofs to be much harder. It is still a win for white though if I let different engines play it out in different configurations (Houdini vs Ivanhoe, Ivanhoe vs Houdini etc.).
The winning scheme is the same as with the pawnending and the minor piece ending, but the black rook has more chances to launch a counter attack, or to cut off your king etc..
The general advice to keep your rook behind your pawns is plain wrong in 95% of the cases.
The knowledge of the theorethical rookendings is useless at this stage of the game.
So far I haven't been able to win this endgame against the computer myself. It offers a unique possibility to get familiar with rook moves. Time trouble in the past always deprived me from that.
I'm going to take my time to learn to play this position with white. A perfect way to learn some endgame technique!
Carlsen the Grinder!
13 hours ago