The PDO/EPB representation is capable of supporting at least some basic functions of a robot reasoning system; the path planning and sliding examples are typical (although simplified) problems in robotics. In solving these problems, the system based on the PDO/EPB representation demonstrates the ability to perform primitive qualitative reasoning about the effects of moving objects in the physical world.
This representation is sufficient to solve the kind of problems that humans expect to be able to solve using qualitative techniques (without measuring workpieces, for instance). Qualitative shape description is sufficient for spatial reasoning to be carried out at the commonsense level at which humans often operate, and it may therefore be sufficient for robot operation to a human-like standard of performance.
The PDO/EPB representation was designed with robot applications in mind, and it incorporates the facilities recommended for robot reasoning in chapter 4. These facilities include reasoning in local contexts (decoupling object relationships in the scene using the proximity transform), multiple levels of detail (in the extended polygon boundary), and relative size description (in the partial distance ordering).
There are of course many limitations of qualitative techniques even when compared to current robot control methods. The loss of accurate positional information would not be acceptable for a robot that has any need to use geometric motion analysis techniques for instance. The methods presented also apply only to two dimensional cases; many three dimensional situations can be reduced to two dimensional motion, but most robots, even at the current level of robot technology, need to plan and operate in three dimensions.