Engines' Ability to Find Forced Checkmates

Take this position for example:

It is white's turn to move here but regardless black's position is crushing (Chess.com engine at depth 18 gives it as -15.0).  Of course humans would have presumably zero chance of finding a forced checkmate or mate in X number of moves here.  But is it currently beyond the scope of even the world's strongest engine to find a such a mate here?  

Black is shortly going to be up a queen, an exchange and a pawn.  White's king will be stuck on its first rank at least for now.  Black's rooks are poised to shortly join into the attack.  Black has a potential passed pawn currently on d4 as well as a passed pawn on e6.  Meanwhile white really has no play whatsoever and can really only play to delay the inevitable defeat.  

Is it possible that someday (if not already) that an engine can be programmed to be strong enough to find a forced checkmate or Mate in X in positions as one sided as the one diagrammed above?

By the way the reason I chose to diagram the position I did rather than the position after the coming Bxf3+ followed by Qxf3+ is only because it is white to move and I was not sure what to play in the interim for white before Bxf3+.  The engine recommends for white to play Bd2 but I am hardly impressed.

Engines are limited by depth and depending on how many moves deep the mate is, may have too many potential positions to reach a concrete mating line for engines to get that far in a reasonable amount of time.

Thanks for replying.  I figured as much. 

I do not pretend to know much about the development of computer technology but is it at all reasonable to think that perhaps one day the calculating process could be sped up enough to allow calculating such mates in maybe say even in a couple of hours a possibility?

I am not the least bit curious how many moves it would take to checkmate after Bd2, Bf4 or any other move because I know mate will come fast enough.  I am solely curious as to whether or not engines have or will have the capability to figure it out.  And it is not even as though I am interested in following such an engine line(s) to see what the mating sequence(s) would be.  

So, finally got a chance to get to a full device to look at this and Stockfish 14.1 NNUE finds

• a mate in 12, at depth 37, if white moves Be3
• a mate in 11 at depth 39 if white moves h4, which changed to mate in 13 after another ply

I'll try running it in another engine later as well and let that run a little longer too. That said, are you actually asking if an engine can find mate in all possible first moves or in what it evaluates to be the best first move? If only the first one, with enough time, it likely can find a line that mates, which should mean that any deviations are also mating, but due to the way engines work by pruning some lines, won't give all forcing variations.

The other thing, related to my question above, is that white has 24 possible first moves. Most of those moves are only going be briefly looked at and get pruned as being worse for white than other options, even though at depth, there may be some resource that postpones mate longer than other options. The permutations at each step are going to prevent giving an absolute forcing line, due to one or more of those permutations and potential pruning missing potential resources.

At depth 44, no mate is found in the top 3 lines for white and the 4th strongest h4, has mate in 11 again (though only at depth 43).

Reran analysis to have it look at all possible first moves, but not letting it get as deep as the first time (since it takes too long), I get this:

I looked at this some more and stepped into the 1. Bf4 and 1. Qd3 lines and Stockfish 14 NNUE found mates in both of those after that one move check (depth 46 and depth 21 respectively) . 1. c4 was the only line that seems to draw things out the longest and still wasn't finding a mate at 38 ply, even after going into the line a few moves. It probably can find one if given a sufficient amount of time, though it likely will take more than an hour additional time.

What's interesting is Dragon 2.6 found the Bf4 line as a mating line a lower depth, but didn't on some of the others up to depth 27

Sorry @Martin_Stahl I did not respond sooner but thank you for all of that extensive analysis with multiple engines at multiple depths.  I was not so interested in the particulars of the provided position.  I just chose that position as an example from a possible sideline of a recently completed game which gave me the idea for this forum. 

However your research was not done in vain.  It does tell me that indeed it is possible (and currently) for an engine to find the minimum number of moves till checkmate in such a one sided position where it is not a matter of if but when mate will be delivered.  

And I would imagine that there are several lines which lead up to mate in the minimum number of moves in positions such as these.  For example moving a rook to a choice of multiple squares on an open file already occupied by that rook.  And probably the higher the minimum number of moves increases the amount of possible mating lines.  So I am actually more interested in the minimum number of moves than the lines because of course the lines themselves can be rather trivial.

Thanks.

Do you mean maximum? Because, depending on the position, the minimum can be can be easy, as long as you force the engine to look at every possible first move, otherwise pruning is is going to throw out short mates because that's not the best defense.

What I was curious about is the least number of moves it takes to checkmate against best defense as opposed to the worst defense.  I apologize for any confusion.

c4 was the only line that didn't find a mate during the time I let it run. It's possible it's there and move pruning and the horizon effect is keeping the mate out of reach in a reasonable time frame.

You are absolutely correct, The Huntsman 1 chess engine says that after White moves that Black can mate in 12 (or less) depending on what White does.

Thanks @drdos7! You have definitely answered my original question!

And thanks to @Optimissed and @Martin_Stahl for your contributions as well.🙂

Engines playing Chess whatever next is that one of those self driving cars, plays chess while it drives you ... whatever next 😆 lol

Bots in chess.com have a big problem, they do not know how to force checkmate with a rook and a king against a lone king. At least the low rated bots do not know how.

A lot of low rated players don't know that.