Closing the Motion Execution Gap: From Semantic Motion Task Constraints to Kinematic Control

arXiv 2026 46.7 method

TLDR

Introduces Motion Statecharts and a differentiable kinematic world model to bridge semantic task constraints and executable robot motions, demonstrated on eight platforms.

Reasoning

The paper presents a novel framework (Giskard) that combines symbolic motion representation with a differentiable kinematic world model, enabling cross-platform transferability. Strengths include open-source availability and real-world deployment on multiple robots. Weaknesses are the focus on kinematic rather than dynamic models, and no mention of learning or generative capabilities.

Read-first score

Read-first score 46.7, weighted from topical fit, citation, graph, method, reproducibility, and recency signals. Original total remains 18.

Recency 6%

100

Uses a gentle age decay so recent papers surface without erasing older foundations. 2026

Reproducibility 18%

Screens links and visible text for paper, code, dataset, artifact, and repository signals. pdf=True; code=True; dataset=False; markers=github,open source

Citation impact 18%

67.2

Uses OpenAlex-shaped citation metadata as a bibliometric attention signal, separate from paper quality. citation_normalized_percentile=0.67203133

Methodology quality 18%

Screens visible abstract and analysis fields for experiment, dataset, baseline, metric, and limitation evidence. markers=none

Topical relevance 29%

25.7

Uses existing LLM keyword relevance scores normalized to 0-100. world model,world simulator,generative world model,interactive world model,video world model,world dynamics prediction,model-based reinforcement learning world model

Citation velocity 12%

Citation velocity estimates citations per publication-year to reduce old-paper bias. velocity=0.00

Field roles

FrontierBridgeReproducibility anchor

Rank sensitivity

Stability: volatile; rank range: 252.

Keyword Scores

world model

interactive world model

world simulator

world dynamics prediction

generative world model

video world model

model-based reinforcement learning world model

Deep Analysis

Innovations

Motion Statecharts as an executable symbolic representation for complex motions, allowing arbitrary arrangement of motion constraints, monitors, or nested statecharts in parallel and sequence.
Unified differentiable kinematic world model of both robots and environments for world-centric motion specification and generalization across embodiments.
lMPC-based implementation of the task-function approach with jerk bounds to ensure smooth transitions during task switches.
Cross-platform transferability demonstrated on eight robot platforms operating in diverse environments, with the open-source framework Giskard.

Methodology

The paper introduces Motion Statecharts, an executable symbolic representation that enables arbitrary arrangement of motion constraints, monitors, and nested statecharts in parallel and sequence. A unified differentiable kinematic world model of robots and environments is used for world-centric motion specification and embodiment generalization. Motion execution is realized via an lMPC-based implementation of the task-function approach, incorporating jerk bounds to ensure smooth transitions during task switches.

Key Results

The framework, called Giskard, was deployed on eight different robot platforms in diverse environments, demonstrating cross-platform transferability and closing the motion execution gap between high-level semantic constraints and executable robot motions.

Resources

Paper Code