Satisficing-Planning MetricTime Track

TPP

In this planning domain there are 40 problems (named by numbers on the x-axis). In the quality plot, the plan metric has to be minimized. For the problems in this domain, the main challenge is computing a valid plan with *good quality*, but for a fully-automated planner in some cases even finding any valid plan is hard. Computing the optimal solution in this domain is known NP-hard problem.








Openstacks

In this planning domain there are 20 problems (named by numbers on the x-axis). In the quality plot, the plan metric has to be minimized. For the problems in this domain, the main challenge is computing a valid plan with *good quality*, but for a fully-automated planner in some cases even finding any valid plan is hard.








Pathways

In this planning domain there are 30 problems (named by numbers on the x-axis). In the quality plot, the plan metric has to be minimized. For the problems in this domain, the main challenge is computing a valid plan with *good quality*, but for a fully-automated planner in many cases even finding any valid plan is hard.




Pipesworld

In this planning domain there are 50 problems (named by numbers on the x-axis). The problems not indicated are solved by no planner. In the quality plot, the plan metric has to be minimized.




Rovers

In this planning domain there are 40 problems (named by numbers on the x-axis). In the quality plot, the plan metric has to be minimized. The problems not indicated are solved by no planner. For the problems in this domain, the main challenge is computing a valid plan with *good quality*, but for a fully-automated planner in some cases even finding any valid plan is hard.




Storage

In this planning domain there are 30 problems (named by numbers on the x-axis). In the quality plot, the plan metric has to be minimized.




Trucks

In this planning domain there are 30 problems (named by numbers on the x-axis). In the quality plot, the plan metric has to be minimized. For the problems in this domain, the main challenge is computing a valid plan with *good quality*, but for a fully-automated planner in some cases even finding any valid plan is hard.