Planning at King's

L-RPG is a domain-independent Forward-Chaining planner which - unlike many other planners - does not rely on grounding for guiding its search through the search space. The ultimate aim of this planner is to be more scalable than planners which rely on grounding and solve bigger problem instances which cannot currently be solved by state-of-the-art planners due to memory constraints.

OPTIC is a temporal planner for use in problems where plan cost is determined by preferences or time-dependent goal-collection costs.  Such problems arise in a range of interesting situations, from scheduling the delivery of perishable goods, to coordinating order-fulfillment activities in warehouses.

POPF is forwards-chaining temporal planner. Its name derives from the fact that it incorporates ideas from partial-order planning — during search, when applying an action to a state, it seeks to introduce only the ordering constraints needed to resolve threats, rather than insisting the new action occurs after all of those already in the plan. Its implementation is built on that for the planner COLIN, and it retains the ability to handle domains with linear continuous numeric effects.

The source code for POPF is available to download from the POPF sourceforge repository. The latest version available, POPF2, corresponds to POPF2 from the 2011 International Planning competition plus fixes for a few bugs that have been found since then.

LPRPG-P is a planner for use with problems containing propositional and numeric preferences.  It uses a combination of a Linear Programme (LP) heuristic to give search guidance in the presence of numeric preferences, with a Relaxed Planning Graph heuristic for propositional preferences.  Click here to download the sources of the IPC version of LPRPG-P.