<s>
Graphplan	B-Application
is	O
an	O
algorithm	O
for	O
automated	B-Application
planning	I-Application
developed	O
by	O
Avrim	O
Blum	O
and	O
Merrick	O
Furst	O
in	O
1995	O
.	O
</s>
<s>
Graphplan	B-Application
takes	O
as	O
input	O
a	O
planning	O
problem	O
expressed	O
in	O
STRIPS	B-Application
and	O
produces	O
,	O
if	O
one	O
is	O
possible	O
,	O
a	O
sequence	O
of	O
operations	O
for	O
reaching	O
a	O
goal	O
state	O
.	O
</s>
<s>
The	O
name	O
graphplan	B-Application
is	O
due	O
to	O
the	O
use	O
of	O
a	O
novel	O
planning	O
graph	O
,	O
to	O
reduce	O
the	O
amount	O
of	O
search	O
needed	O
to	O
find	O
the	O
solution	O
from	O
straightforward	O
exploration	O
of	O
the	O
state	O
space	O
graph	O
.	O
</s>
<s>
On	O
the	O
contrary	O
,	O
in	O
Graphplan	B-Application
's	O
planning	O
graph	O
:	O
</s>
<s>
The	O
algorithm	O
then	O
iteratively	O
extends	O
the	O
planning	O
graph	O
,	O
proving	O
that	O
there	O
are	O
no	O
solutions	O
of	O
length	O
l-1	O
before	O
looking	O
for	O
plans	O
of	O
length	O
l	O
by	O
backward	O
chaining	O
:	O
supposing	O
the	O
goals	O
are	O
true	O
,	O
Graphplan	B-Application
looks	O
for	O
the	O
actions	O
and	O
previous	O
states	O
from	O
which	O
the	O
goals	O
can	O
be	O
reached	O
,	O
pruning	O
as	O
many	O
of	O
them	O
as	O
possible	O
thanks	O
to	O
incompatibility	O
information	O
.	O
</s>
<s>
A	O
closely	O
related	O
approach	O
to	O
planning	O
is	O
the	O
Planning	O
as	O
Satisfiability	O
(	O
Satplan	B-Application
)	O
.	O
</s>
<s>
Both	O
reduce	O
the	O
automated	B-Application
planning	I-Application
problem	O
to	O
search	O
for	O
plans	O
of	O
different	O
fixed	O
horizon	O
lengths	O
.	O
</s>
