Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Motion planning with complete knowledge using a colored SOM.

J Vleugels1, J N Kok, M Overmars

  • 1Department of Computer Science, Utrecht University, The Netherlands.

International Journal of Neural Systems
|March 5, 1999
PubMed
Summary

This study demonstrates neural networks effectively solve robot motion planning with complete configuration knowledge. A novel approach combines neural networks and deterministic methods to create a motion roadmap, outperforming traditional techniques.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

New Speed-Sintered CaO-Zirconia for Strong, Tough, Translucent Restorations.

Journal of dental research·2026
Same author

Coloring Multilayer Zirconia May Affect Its Optical and Mechanical Properties.

Journal of dental research·2024
Same author

Notch ratio in pulmonary flow predicts long-term survival after pulmonary endarterectomy for chronic thromboembolic pulmonary hypertension.

Heart and vessels·2024
Same author

Grain-Boundary Engineering for Aging and Slow-Crack-Growth Resistant Zirconia.

Journal of dental research·2017
Same author

Controlled release of chlorhexidine from a mesoporous silica-containing macroporous titanium dental implant prevents microbial biofilm formation.

European cells & materials·2017
Same author

Detecting dispersed duplications in high-throughput sequencing data using a database-free approach.

Bioinformatics (Oxford, England)·2015

Area of Science:

  • Robotics
  • Artificial Intelligence
  • Computational Geometry

Background:

  • Robot motion planning is crucial for autonomous systems navigating complex environments.
  • Existing neural network methods often rely on limited, local environmental information.
  • Complete knowledge of the configuration space presents unique challenges for pathfinding algorithms.

Purpose of the Study:

  • To demonstrate the efficacy of neural networks for robot motion planning using complete configuration space knowledge.
  • To introduce a hybrid approach combining neural networks with deterministic algorithms.
  • To develop a robust roadmap generation method that overcomes limitations of existing techniques.

Main Methods:

  • A novel colored Kohonen's self-organizing map with two node classes was developed.

Related Experiment Videos

  • The network learns from random robot configurations to build a motion roadmap.
  • Obstacle boundaries and Voronoi diagrams are approximated by distinct node types.

    • Main Results:

    • The hybrid approach successfully constructs a motion roadmap from complete configuration data.
    • The method efficiently handles complex scenes with small obstacles and narrow passages.
    • Experimental results show superior performance compared to conventional motion planning techniques.

    Conclusions:

    • Neural networks are viable tools for robot motion planning with global environmental awareness.
    • The proposed hybrid method offers a simple, general, and effective solution for pathfinding.
    • The approach demonstrates strong performance in challenging planar robot motion planning scenarios.