Software

WiSM: Windowing Surrogate Model for Fast Evalution of Curvature-Constrained Dubins Tours

Here, link to source codes of the WiSM approximator will be placed in a case of the corresponding paper would be accepted.

Fast Heuristics for GTSPN-based 3D Multi-Goal Path Planning

Here, link to source codes of the fast heuristics solver for the GTSPN will be placed in a case of the corresponding paper would be accepted.

 

Adaptive Locomotion Control of Hexapod Robot with Position Feedback Only in Rough Terrains

Here, link to source codes of the adaptive locomotion controller will be placed in a case of the corresponding paper would be accepted.

 

Growing Self-Organizing Array (GSOA) for Eucliead TSP and Close-Enough TSP (CETSP)

A public release of the GSOA implementation for solving Euclidean Traveling Salesman Problem (ETSP) and Close Enough TSP (CETSP) including problem instances used in the paper are provided at the https://github.com/comrob/gsoa repository. Further details about the GSOA can be found in the following paper.

  • [DOI] J. Faigl, “GSOA: growing self-organizing array – unsupervised learning for the close-enough traveling salesman problem and other routing problems,” Neurocomputing, vol. 312, pp. 120-134, 2018.
    [Bibtex]
    @article{faigl18gsoa,
    author = {Jan Faigl},
    title = {{GSOA:} Growing Self-Organizing Array - Unsupervised learning for
    the Close-Enough Traveling Salesman Problem and other routing problems},
    journal = {Neurocomputing},
    volume = {312},
    pages = {120--134},
    year = {2018},
    url = {https://doi.org/10.1016/j.neucom.2018.05.079},
    doi = {10.1016/j.neucom.2018.05.079},
    }

Surveillance Planning with Bézier Curves

Public release of the GSOA-based approach (formerly SOM-based) for solving surveillance planning problem formulated as a variant of the Close Enough Traveling Salesman Problem (CETSP) with Bézier curves, i.e., a variant of the TSP with Neighborhoods (TSPN) with disk-shaped neighborhoods in 2D and sphere-shaped neighborhoods in 3D, is provided at the https://github.com/comrob/cetsp-bezier-gsoa. The approach has been introduced in our RA-L paper and then further generalized to solve problems with a group of aerial vehicles reported in the Journal of Field Robotics. Notice, the employed unsupervised learning based approach is originally based on the Self-Organizing Map (SOM) for the TSP that has been consolidated and generalized to the Growing-Self-Organizing Array (GSOA) for solving TSP like routing problems with the available implementation for solving instances of the Euclidean TSP and CETSP.

  • [DOI] J. Faigl and P. Vana, “Surveillance planning with bézier curves,” IEEE robotics and automation letters, vol. 3, iss. 2, pp. 750-757, 2018.
    [Bibtex]
    @article{faigl18ral,
    author = {Jan Faigl and Petr Vana},
    title = {Surveillance Planning With B{\'{e}}zier Curves},
    journal = {{IEEE} Robotics and Automation Letters},
    volume = {3},
    number = {2},
    pages = {750--757},
    year = {2018},
    url = {https://doi.org/10.1109/LRA.2018.2789844},
    doi = {10.1109/LRA.2018.2789844},
    }
  • J. Faigl, P. Vana, R. Penicka, and M. Saska, “Unsupervised learning based flexible framework for surveillance planning with aerial vehicles,” Journal of field robotics.
    [Bibtex]
    @article{faigl18jfr,
    author = {Jan Faigl and Petr Vana and Robert Penicka and Martin Saska},
    title = {Unsupervised Learning based Flexible Framework for Surveillance Planning with Aerial Vehicles},
    journal = {Journal of Field Robotics},
    note = {In press},
    }

Self-Organizing Map based solver for the Prize-Collecting Traveling Salesman Problem with Neighborhoods (PC-TSPN)

A public release of library for solving PC-TSPN for academic purposes is available in the archive som_pctspn_v1.1.tar.xz. It contains a complete configurable program with visualization. It depends on the following libraries: log4cxx, boost, and cairo. If you use the solver for your research, please, cite the paper.

A public release of the simplified the library without dependencies for solving the PC-TSPN for academic purposes is planned to be published at the end of 2015 (also accompanied with the data set used in our papers). If you have any question regarding the solver do not hesitate to contact me at faiglj@fel.cvut.cz

A visualization of the unsupervised learning during solving an instance of the PC-TSPN can be see in the video file.
Further details about the solver can be found in the following paper.

  • [DOI] J. Faigl and G. A. Hollinger, “Unifying multi-goal path planning for autonomous data collection,” in IEEE/RSJ international conference on intelligent robots and systems, 2014, pp. 2937-2942.
    [Bibtex]
    @inproceedings{faigl14iros,
    author = {Jan Faigl and Geoffrey A. Hollinger},
    title = {Unifying multi-goal path planning for autonomous data collection},
    booktitle = {{IEEE/RSJ} International Conference on Intelligent Robots and Systems},
    pages = {2937--2942},
    year = {2014},
    doi = {10.1109/IROS.2014.6942967},
    }

Frontier-Based Multi-Robot Exploration

A public release of software library for evaluation of the multi-robot exploration strategies in frontier-based grid map exploration is available for academic purposes as an archive mre-ver0.9.tar.xz or at the git repository.
If you have any question regarding the framework do not hesitate to contact me at faiglj@fel.cvut.cz

  • J. Faigl and M. Kulich, “On benchmarking of frontier-based multi-robot exploration strategies,” in Proceedings of 7th European Conference on Mobile Robots, Lincoln, 2015.
    [Bibtex]
    @inproceedings{faigl15ecmr,
    author={Faigl, J. and Kulich, M.},
    title={On Benchmarking of Frontier-Based Multi-Robot Exploration Strategies},
    booktitle = {{Proceedings of 7th European Conference on Mobile Robots}},
    year={2015},
    address={Lincoln},
    notes={(to appear)}
    }