Learning Human Preferences to Inform Soft Constraints for Robot Planning

The core message of this work is that human preferences over robot behavior can be effectively learned and represented as soft constraints in a planning framework, enabling adaptable and personalized robot behavior.