Keypoint-based Diffusion for Robotic Motion Planning on the NICOL Robot

2025-09-04

AI Summary

Traditional numerical motion planning methods suffer from excessive computational latency, hindering real-time deployment. To address this, we propose a generative pose model based on keypoint-guided diffusion. Departing from conventional point-cloud encoding as conditional input, our approach models the high-dimensional state space via a keypoint diffusion mechanism, circumventing intricate geometric reasoning. The model jointly learns keypoint prediction and joint trajectory generation under supervision from offline planning data. Evaluated on standard benchmarks, it achieves a 90% collision-free path success rate while reducing planning time by an order of magnitude compared to classical planners. This work introduces the first keypoint-guided diffusion architecture for robotic motion planning, significantly advancing the efficiency and practicality of deep learning–based planners under stringent success-rate constraints.

Technology Category

• Intelligent Robots: Motion and Path Planning
• Planning, Routing, and Scheduling: Planning with Language Models
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

We propose a novel diffusion-based action model for robotic motion planning. Commonly, established numerical planning approaches are used to solve general motion planning problems, but have significant runtime requirements. By leveraging the power of deep learning, we are able to achieve good results in a much smaller runtime by learning from a dataset generated by these planners. While our initial model uses point cloud embeddings in the input to predict keypoint-based joint sequences in its output, we observed in our ablation study that it remained challenging to condition the network on the point cloud embeddings. We identified some biases in our dataset and refined it, which improved the model's performance. Our model, even without the use of the point cloud encodings, outperforms numerical models by an order of magnitude regarding the runtime, while reaching a success rate of up to 90% of collision free solutions on the test set.

Problem

• Proposing diffusion-based action model for robotic motion planning
• Reducing runtime requirements compared to numerical planning approaches
• Achieving high success rate for collision-free motion solutions

Innovation

• Keypoint-based diffusion model for motion planning
• Leveraging deep learning to reduce runtime
• Refined dataset to improve model performance