Graph-based Planners

Graph-based Planners are based on ideas used in graph algorithms. Given a problem statement, a Graph-based Planner explicitly constructs and annotates a compact structure called a Planning Graph, in which a plan is a kind of "flow" of truth-values through the graph. This graph has the property that useful information for constraining search can quickly be propagated through the graph as it is being built. The Graph-based Planner then exploits this information in the search for a plan[1].

Graphplan: The first planner to use a planning graph, developed by Avrim Blum and Merrick Furst from the Carnegie Mellon University.

Blackbox: blackbox = satplan + graphplan, developed by Henry Kautz at the University of Washington.

DPPLAN : A planner based on the planning graph, using different search procedures on the planning graph, which are not compelled to be directional. It has been developed by Marco Baioletti with the help of Alfredo Milani from the Universities of Perugia and Siena.

IPP : The IPP planning system was developed by Jana Koehler, Joerg Hoffmannr, Michael Brenner, and Frank Rittinger at the University of Freiburg from Fall 1996 to Spring 1999.

LPG developed by Alfonso Gerevini and Ivan Serina from the University of Brescia.

PropPlan from Michael Fourman at the University of Edinburgh.

Sensory Graphplan developed by Dan Weld et al from the University of Washington.

STAN developed by Maria Fox and Derek Long at the University of Strathclyde.

TGP Temporal Graphplan, from Dan Weld et al at the University of Washington.


Print Page     PLANET Home