Deep Affordance Foresight: Planning Through What Can Be Done in the Future
This work addresses the challenge of efficient planning in complex environments for robotics, offering a novel approach to affordance-based reasoning that is not incremental but introduces a new paradigm for long-term task achievement.
The paper tackles the problem of long-horizon planning in robotics by introducing a new affordance representation that models future action outcomes, enabling robots to select actions for achieving multi-step tasks. The result is Deep Affordance Foresight (DAF), which effectively learns to perform tasks in challenging manipulation domains, as demonstrated through evaluation on two domains with capabilities like sharing representations and handling high-dimensional inputs.
Planning in realistic environments requires searching in large planning spaces. Affordances are a powerful concept to simplify this search, because they model what actions can be successful in a given situation. However, the classical notion of affordance is not suitable for long horizon planning because it only informs the robot about the immediate outcome of actions instead of what actions are best for achieving a long-term goal. In this paper, we introduce a new affordance representation that enables the robot to reason about the long-term effects of actions through modeling what actions are afforded in the future, thereby informing the robot the best actions to take next to achieve a task goal. Based on the new representation, we develop a learning-to-plan method, Deep Affordance Foresight (DAF), that learns partial environment models of affordances of parameterized motor skills through trial-and-error. We evaluate DAF on two challenging manipulation domains and show that it can effectively learn to carry out multi-step tasks, share learned affordance representations among different tasks, and learn to plan with high-dimensional image inputs. Additional material is available at https://sites.google.com/stanford.edu/daf