There are a number of sub-problems within the path planning problem which could be solved by the methods used for the sliding problem. For instance, the safest route through a narrow gap is to slide along one side of the gap.
Although the sliding system was implemented before I developed the EPB/PDO representation, the methods used would be easily adapted to this representation. In fact, an implementation using EPB/PDO would be simpler than the one using ASSF. If this was done, it would not be difficult to integrate the two planning methods into a system that could choose between different motion strategies during path planning.
The main characteristics of these two strategies are boundary following for the sliding system, and straight line motion between a series of via points (such as the narrowest point in a gap, and the entrance to a gap) in the path planning system. A framework which could direct either of these strategies could also include further strategies, such as rotation, wiggling into place, or other motion strategies regularly used by people. The addition of new motion strategies to an overall task planning structure provides an interesting basis for comparison to human acquisition of manipulation skills, as shown in Sussman's HACKER system [Sus75].