Example

Example: Putting on a pair of shoes

$Init()$

$Goal(RightShoeOn \wedge LeftShoeOn)$

$Action(\textcolor{pink}{RightShoe}, \text{PRECOND: }RightSockOn, \text{EFFECT: }RightShoeOn)$

$Action(\textcolor{pink}{RightSock}, \text{EFFECT: }RightSockOn)$

$Action(\textcolor{pink}{LeftShoe}, \text{PRECOND: }LeftSockOn, \text{EFFECT: }LeftShoeOn)$

$Action(\textcolor{pink}{LeftSock}, \text{EFFECT: }LeftSockOn)$

Partial-order planner would come up seperately with:

$[RightSock, RightShoe]$ for goal $RightShoeOn$

$[LeftSock, LeftShoe]$ for goal $LeftShoeOn$

Then these sequences would be combined to construct a final plan.

Every action sequence that maintains the partial order is a solution.

"Every linearisation of a partial-order solution is a total-order solution whose execution from the initial state will reach a goal state."

Solution by POP

Actions

Orderings

Links

Open preconditions