|
Director: Julien Bourgeois (julien.bourgeois@femto-st.fr)
Supervisor: Benoît Piranda (benoit.piranda@femto-st.fr)
This proposal is part of an ongoing project that aims to realize a self-reconfigurable programmable matter in 3D, i.e. the spatial self-reorganization of micro-robots so that they reach a goal configuration defined in advance.
Known as modular robotics, this matter aims to provide options and capabilities in order to change one of its physical proprieties like shape or locomotion in reaction to internal or external action [1]. Many works propose distributed self-reconfiguration algorithms in 2D like in [2] or [4], but self-reconfiguration of a set of robots in a 3D space remains a great challenge of the domain. There is no efficient solution in literature that is adapted to the fully distributed context of programmable matter.
Some recent works of our team proposed the design of a quasi-spherical modular robot capable of moving relatively to its neighbors by successive rotations [1]. These micro-robots called 3D Catoms are placed in a Face Centered Cubic lattice (FCC) giving, up to 12 neighbors per robots. This high density is an asset for multiplying possible displacements and optimizing communications [3].
In the context of this PhD, we will focus on the definition of specific distributed self-reconfiguration algorithms that take into account physical laws such as forces generated by gravity or the elastic membrane. In particular, the physical stability of all the intermediate configurations produced to achieve the goal configuration that must be guaranteed by the algorithm.
For several years, our laboratory has been developing and operating a behavioral simulator to visualize modular robots in their virtual environment (VisibleSim [2]). The physical models developed during the thesis must be integrated into our simulator. This powerful tool will allow to validate self-reconfiguration algorithms for large sets of connected robots.
To summarize, the successful candidate will participate to the realization of one of the first fully distributed 3D self-reconfiguration algorithm for programmable matter.
Candidate profile:
The candidate should have a major in computer science. A background in physical simulation would be appreciated.
References:
[1] Benoit Piranda and Julien Bourgeois. Geometrical Study of a Quasi-Spherical Module for Building Programmable Matter. 13th International Symposium on Distributed Autonomous Robotic Systems (DARS-2016), In press, 2016.
[2] Benoit Piranda, Guillaume J. Laurent, Julien Bourgeois, Cédric Clévy, and Nadine Le Fort-Piat. A new concept of planar self-reconfigurable modular robot for conveying microparts. Mechatronics, 23(7):906–915, October 2013.
[3] André Naz, Benoît Piranda, ThadeuKnychalaTucci, Seth Copen Goldstein and Julien Bourgeois, “Network Characterization of Lattice-Based Modular Robots with Neighbor-to-Neighbor Communications”, 13th International Symposium on Distributed Autonomous Robotic Systems (DARS-2016), In pres, 2016
[4] M. Yim, D. G. Duff, and K. D. Roufas, “Polybot: a modular reconfigurable robot,” in IEEE International Conference on Robotics and Automation (ICRA), vol. 1, 2000, pp. 514–520.
|
