Classical Planning in Deep Latent Space
This addresses the problem of integrating deep learning with symbolic planning for researchers in AI and robotics, though it is incremental as it builds on existing methods.
The paper tackles the knowledge acquisition bottleneck in classical planning by proposing Latplan, an unsupervised architecture that learns a complete propositional PDDL action model from unlabeled image pairs, enabling planning in a symbolic latent space and achieving successful plan execution across six image-based planning domains.
Current domain-independent, classical planners require symbolic models of the problem domain and instance as input, resulting in a knowledge acquisition bottleneck. Meanwhile, although deep learning has achieved significant success in many fields, the knowledge is encoded in a subsymbolic representation which is incompatible with symbolic systems such as planners. We propose Latplan, an unsupervised architecture combining deep learning and classical planning. Given only an unlabeled set of image pairs showing a subset of transitions allowed in the environment (training inputs), Latplan learns a complete propositional PDDL action model of the environment. Later, when a pair of images representing the initial and the goal states (planning inputs) is given, Latplan finds a plan to the goal state in a symbolic latent space and returns a visualized plan execution. We evaluate Latplan using image-based versions of 6 planning domains: 8-puzzle, 15-Puzzle, Blocksworld, Sokoban and Two variations of LightsOut.