Learning Sequential Kinematic Models from Demonstrations for Multi-Jointed Articulated Objects

TL;DR

This paper introduces a new method for learning models of multi-jointed objects from human demonstrations, enabling robots to better understand and manipulate complex objects with multiple degrees of freedom, even when some joints are occluded.

Contribution

The paper presents OKSMs, a novel kinematic sequence model, and Pokenet, a neural network for estimating object models from point clouds, advancing manipulation capabilities for multi-DoF objects.

Findings

Pokenet improves joint estimation accuracy by over 20% on real data.

OKSMs successfully model manipulation sequences for multi-DoF objects.

Demonstrated manipulation of complex objects on a Sawyer robot.

Abstract

As robots become more generalized and deployed in diverse environments, they must interact with complex objects, many with multiple independent joints or degrees of freedom (DoF) requiring precise control. A common strategy is object modeling, where compact state-space models are learned from real-world observations and paired with classical planning. However, existing methods often rely on prior knowledge or focus on single-DoF objects, limiting their applicability. They also fail to handle occluded joints and ignore the manipulation sequences needed to access them. We address this by learning object models from human demonstrations. We introduce Object Kinematic Sequence Machines (OKSMs), a novel representation capturing both kinematic constraints and manipulation order for multi-DoF objects. To estimate these models from point cloud data, we present Pokenet, a deep neural network trained on human demonstrations. We validate our approach on 8,000 simulated and 1,600 real-world annotated samples. Pokenet improves joint axis and state estimation by over 20 percent on real-world data compared to prior methods. Finally, we demonstrate OKSMs on a Sawyer robot using inverse kinematics-based planning to manipulate multi-DoF objects.