Previous MEETROBOT meetings

From IridiaWiki
Jump to navigationJump to search

Wednesday 10th May 2006 - Shervin

Agenda

  1. Presentation / Discussion
    • I am going to present an extension of my chaining controller. In the original version robots just form chains from the nest. In the new controller I introduce a very simple "prey enhancement" mechanism where the robots activate their LEDs when they perceive the nest, and in this way enhance the area in which the prey can be perceived. I present the results from an experiment I conducted in simulation.
  2. Robotics Group Admin

Reading material


People who will be absent





Wednesday 3rd May 2006 - Roderich

Agenda

  1. Presentation / Discussion
    • I'm going to rehearse two presentations that have to be prepared for the 2006 IEEE Int. Conf. on Robotics & Automation (ICRA), which is held in two weeks in Florida. Each presentation should last 13 minutes + discussion.
    • Abstract: This paper addresses the cooperative transport of a heavy object by a group of mobile robots. We present a system in which group members lacking knowledge about the position of the transport target exploit physical interactions with other members of the group that have such knowledge. This is the first such system to achieve a performance superior to that of a passive caster. The system is fully decentralized and the information flow between the robots is limited to physical interactions. The robots have no knowledge about their relative positions. A comprehensive experimental study with up to six physical robots confirms the effectiveness, reliability, and robustness of the system. Finally, the system is examined in rough terrain conditions.
    • Abstract: We present a first attempt to accomplish a simple object manipulation task using the self-reconfigurable robotic system swarm-bot. The number of modular entities involved, their global shape or size and their internal structure are not pre-determined, but result from a self-organized process in which the modules autonomously grasp each other and/or an object. The modules are autonomous in perception, control, action, and power. We present quantitative results, obtained with six physical modules, that confirm the utility of self-assembling robots in a concrete task.
  1. Robotics Group Admin

Reading material

These are the two papers. However, the presentations should be self-contained.

http://iridia.ulb.ac.be/~rgross/icra06_568_final.pdf

http://iridia.ulb.ac.be/~rgross/icra06_922_final.pdf


People who will be absent


Wednesday 26th April 2006 - Christos

I will talk about current work on the evolution of communication for heterogeneous robots, trying to link it with older work and also with what I have in mind as future goal, but also on a new self-assembly method we are working on with Anders, where the NN takes care of everything, without use of camera. I will also elaborate on the doors the latter effort might open once successful.

Agenda

  1. Presentation / Discussion
  2. Robotics Group Admin

Reading material


People who will be absent


Results




Wednesday 19th April 2006 - Elio

Agenda

  1. Presentation / Discussion
    • I presents the results of a series of simulation concerning the evolution of controllers for heterogeneous robots required to coordinate their actions in order to approach a target. This work, carried out by myself, Christos Ampatzis, Federico Vicentini and Marco Dorigo, has been recently submitted to SAB06.
  1. Robotics Group Admin

Reading material If you are interested in doing some reading about this subject, you can find in my room printed copies of the paper submitted to the conference.

People who will be absent





Wednesday 13th April 2006 - Anders

Agenda

  1. Presentation / Discussion
    • Controller design
    • Palet lifting and transport
    • Evolving self-assembly
  2. Robotics Group Admin
    • E-pucks

Reading material

None

People who will be absent

  • MarcoD, Rehan




Wednesday 6th April 2006 - Rehan

Agenda

  1. Presentation / Discussion
    • Controller - Two robots adaptively rotate to overcome a hill
  2. Robotics Group Admin
    • Rehan - Common Interface Wiki Page

Reading material

  • None

People who will be absent

  • MarcoD




Wednesday 29th March 2006 - Alexandre

Agenda

  1. Presentation / Discussion
    • The presentation is about the results of my experiments on negotiation of direction for cooperative transport.
    • I'll explain the experiments, present and discuss the results, and also ask your opinion about future work.
    • The (say) ten last minutes I'll present you the next work I plan to do.

Open document presentation (openoffice2) available : http://iridia.ulb.ac.be/~acampo/various/nego021.odp The same as a pdf : http://iridia.ulb.ac.be/~acampo/various/nego021.pdf

  1. Robotics Group Admin

Reading material

Take a look at the paper submitted to ANTS2006, it summarizes the results obtained in the negotiation experiment. (has been removed now)


People who will be absent

  • Christos, MarcoD





Wednesday 1st March - Shervin

Agenda

  1. Presentation / Discussion
  • Shervin would like to rehearse the talk he will give next week at the IAS-9 conference. The talk is very similar to the one given in October, so for those who already followed it by then it might be quite boring to hear the same stuff again....
  1. Robotics Group Admin

Reading material


People who will be absent

Marco D, Thomas


Results





Wednesday 22nd February - Roderich

Agenda

  1. Presentation / Discussion
  • Roderich would like to (prepare and) rehearse his talk "Self-assembly of Mobile Robots - From Swarm-bot to Super-mechano Colony" to be held on March 7, 2006 at the 9th Int. Conf. on Intelligent Autonomous Systems, IAS-9, Tokyo.

The talk will be quick and can be followed without reading the paper (see link).

http://iridia.ulb.ac.be/~rgross/reading_material.pdf

  1. Robotics Group Admin

Reading material


People who will be absent

Marco D, Thomas

Results




Wednesday 15th February - Marco

Agenda

  1. Presentation / Discussion
  • Marco: I will briefly present the Swarmanoid project: goals, structure and allocated resources (15 minutes max, everybody invited, also optimization people, so that everybody is aware of Swarmanoid).
  1. Robotics Group Admin

Reading material


People who will be absent

Results





Wednesday 8th February - Christos

Agenda

  1. Presentation / Discussion
    • Christos: Given that I do not have something completely new to present, I will discuss 2-3 intresting points that came out of the work presented at the previous talk. I will also complement Elio's talk discussing a possible interesting scenario. Some ideas for my future work will also be given.
  2. Robotics Group Admin
    • Anders: Fixed-point: How to handle precision arithmetic on the s-bot without the use of floats/doubles.
    • Rodi and Anders: Cooperative transport of an object on a pallet, status and ideas.
    • Alex: proudly announces that it is possible to detect reliably a direction pointed by a robot using visual communication (LED-camera)

Reading material


People who will be absent

Elio


Results




Wednesday 1st Febuary - Elio

Agenda

  1. Presentation / Discussion

This week I am going to give a "VERY INFORMAL" (please note the "informal") presentation on some work I'm currently doing on the evolution of signaling using Evolutionary Robotics methods. Note that, the meeting will be strictly focused on the description of my experiments (which, unfortunately for me, they are not very successful at the moment). So, if you are not interested in listening my sorrowful talk, you can certainly miss it.

  1. Robotics Group Admin

Reading material


People who will be absent

Results





Wednesday 25th January - Anders

This meeting is going to be a bit technical.

Agenda

  1. Presentation / Discussion
    • Simulation
    • TwoDee: Philosophy, limitations and implementation
    • Discussion on simulation
  2. Robotics Group Admin

Reading material

No reading is required, but I would recommend that people interested in or working with simulation to read Jacobi's "Evolutionary Robotics and the Radical Envelope-of-Noise Hypothesis" (.ps.Z), (1997), Journal of Adaptive Behavior vol. 6, No. 2, 325-368.

People who will be absent Rehan

Results

Download presentation (.pdf)




Wednesday 18th January - Rehan

Agenda

  1. Presentation / Discussion
  • Sbot Common Interface
    • Design Principles Design
    • Implementation (Current)
    • Live Demonstration
  1. Robotics Group Admin

Reading material


People who will be absent

Results

Download the presentation (.pdf)


Wednesday 11th January - Alexandre

Agenda

  1. Presentation / Discussion
  • I will introduce a new experiment that involves a group of robots trying to estimate collectively

the direction of the nest, whereas they can not see it anymore.

  • After a quick description of the experiment I'll introduce a set of simple behavioural rules devised to produce the desired collective behaviour.
  • Then I'll present some simulation results, and describe how the group of robots manages to estimate the direction

of the nest, and how it behaves in varying conditions (noise of the system, number of robots, duration of the experiment...).

  • Lastly I will discuss some tradeoff problems : robots estimate better the direction

of the nest if they are numerous, but on the other hand they have difficulties to synchronize for achieving a collective prey retrieval.

  • And I will discuss some scalability and bifurcation features of the system
  • If I have some time, but it is not likely, I will present you my latest results on optimal foraging

and improvements I did, following the suggestions of many of you.


  1. Robotics Group Admin
    • Anders: Robotics room server
    • Anders: s-bot status (Rehan is going to kidnap the robots?)
    • Anders: BibTeX file for robotics conferences and journals.
    • Marco: Who takes care of the e-pucks?.
    • Marco: NASA wants to cooperate with us.


Reading material


People who will be absent

Results

Download the presentation (.pdf)

  • Anders: Robotics room server
    • There is a new computer in the arena. It needs to be installed and configured. For people who do not have laptops and for the overhead camera, etc.
  • Anders: s-bot status (Rehan is going to kidnap the robots?)
    • Rehan is (maybe) going to take all the s-bots to Switzerland the final week of January.
  • Anders: BibTeX file for robotics conferences and journals.
    • A common BibTeX DB is already on the way. In a months time we should be able to test it.
  • Marco: Who takes care of the e-pucks?.
    • The e-pucks will arrive soon. Noone volunteered to test them. Marco will pick one at random.
  • Marco: NASA wants to cooperate with us.
    • IRIDIA is likely going to sign a contract with NASA for a number of experiments, in simulation and on the real robots, for transporting (lifting) building material on the moon.




Wednesday 14th December - Shervin

Agenda

  1. Presentation / Discussion
    • There's not too much that changed since my last talk.
    • I will (very) shortly present the reimplemented version of my controller on twodee,
    • talk about the experiments which I prepared and which I intend to run before xmas,
    • introduce the idea for a new controller.
    • If there's any interest I can show photos from Cuba....
  2. Robotics Group Admin
    • The last meeting before Christmas (Dec 21st)
    • MEETROBOT 2006


Reading material


People who will be absent


Results

Download presentation (pdf)





Wednesday 7th December - Movie

Agenda

  1. Presentation / Discussion
  2. Robotics Group Admin


Reading material


People who will be absent

Marco: I am currently flying from Lyon to Dublin




Wednesday 30th November - Roderich

Agenda

  1. Urgent stuff
    • MARCO: We need to briefly discuss the ongoing reparation of our s-bots


  1. Presentation / Discussion
    • This talk is related to my visit to Japan. I will provide an overview about:
    • reasons to go there
    • how to get funding (JSPS programs)
    • some robotics research labs
    • my research and the environment
  2. Robotics Group Admin


Reading material

  • NONE


People who will be absent


Results

Download presentation (pdf ~40MB)

  • For the real s-bots the batteries are going to be replaced.
  • The common interface is ready. It currently works on the real s-bots and in the TwoDee simulator.




Wednesday 23rd November - Christos

Agenda

  1. Presentation / Discussion
    • Communication - some brief overview, some definitions (this was supposed to have been discussed some months ago when nobody showed up to discuss a related paper I proposed, I guess this time with some "pressure" you will all be there and super-motivated and interested, what a wonderful world...
    • I am going to talk on my work on the Evolution of Communication in cases where there is no fitness reward for signalling-reaction to the signal. What conditions seem to favour the evolution of communication?
    • An idea for extention, an intro for my next talk, maybe?
  2. Robotics Group Admin
    • Arena camera


Reading material


People who will be absent

Ciro (holiday --> no cake, but maybe a havana cigar...)

Results

Download the presentation (pdf)

  • No news on the arena camera
  • Rehan and Anders are still working (hard) on the common s-bot interface.


Wednesday 16th November - Elio

Agenda

  1. Presentation / Discussion
    • Title: General, Universal, even worst Global: I'm against, better ANTI-.

The talk is meant to be a brief introduction mixed with some personal conclusions concerning the Theory of Direct Perception and its significance in the domain of Evolutionary Robotics.

Robotics Group Admin

  • Arena camera
  • Common interface
  • OMiss and 2D kinematics
  • Duration of the meetings


Reading material

  • None for the moment, I am happy to suggest you some readings if you feel interested in what I will talk about.


People who will be absent

Marco, Federico, Rodi

Results

Download the presentation (pdf)

  • Arena camera: Federico was not present - postponed to next meeting
  • Common interface: Rehan and Anders are working on it now.
  • OMiss and 2D kinematics: It is a good idea, but no one at IRIDIA really needs it now, and no one has the time to work on it.
  • Duration of the meetings: We will try to limit the meetings to 60 mins.



Wednesday 9th November - Federico

Agenda

  1. Presentation / Discussion
    • ER approach for real world apllication: reference framework for vehicle controller
    • Some requirements in my navigation task: patrolling, collision avoidance, obstacle avoidance
    • Multirobot context: testing the Quinn experiment. Simulation and results with real s-bots
    • Discussion/feedback on tools/approaches selected and on future work
  2. Robotics Group Admin
    • Arena camera
    • Common interface
    • OMiss and 2D kinematics


Reading material

  1. Reference for my simulator benchmarking: the Quinn experiment @ Sussex (pdf)
  2. Reference for automotive navigation usign evolutionary approach (pdf)


People who will be absent

N/A


Results

Download the presentation (pdf)

  • We had no time to discuss the points on the agenda, so they have been postponed till next meeting.



Wednesday 2nd November - Rehan

Agenda

  1. IRIDIA Admin
    • Coffee Machine
    • New Wiki Pages
    • Next week...
  2. Presentation / Discussion
    • What is robotic pattern formation?
    • Potential applications using the Swarm Bot robotic platform
    • Discussion / feedback on my research direction
  3. Robotics Group Admin
    • Sbot / Simulator programmatic interfaces
    • Buying a computer+flatscreen for the robot room (or taking one that is not used)
    • Any spare monitor for Rodi?


Location Ping pong room


Reading Material

  • There is no reading material for this weeks presentation


People who will be absent

Elio, Thomas

Results

Download Presentation (PDF)

  • Coffee machine: Various options available. Ciro is taking care of calling the people in charge of the ULB expresso machine deal.
  • New wiki pages:
  • Career developement plans online: Alex is going to propose a technology for the next meeting, which allows us to have up-to-date career development plans online, but only accessible by the student in question and his/her supervisor.
  • Arena camera: Federico is going to look into camera software and hardware for next weeks meeting (for the arena, since the movies shot with the current camera are of poor quality.
  • OMiss with 2D dynamics: We all believe that having the OMiss work with a 2D dynamics engine as well as ODE would be a good idea for many experiments. It might take too much time to implement though and no one seemed too keen on actually implementing it. We will see at the next meeting.


Wednesday 26th October - Anders

Presentation / Discussion Topics

Dependability and fault-tolerance in Robotics:

  • What is fault-tolerance and dependability?
  • What has been done in robotics?
  • How could fault tolerance be implemented in a swarm robotics context?

Administration Issues

  • Coffee machine
  • New documents on the Wiki.
  • Next week...
  • The new e-pucks
  • Simulators and swarm-bot interfaces

People who will be absent

Elio, Rodi, Christos

Results

Download Presentation(PDF)

  • General: The topics on the agenda, which are of general interest are moved to the beginning of the meeting, so that non-robotics people can attend at a fixed hour each week.
  • Coffee machine:
    • Marco buys a new one when he comes across a good, new coffee machine (suggestions are very welcome)
    • Mauro is in charge of sending the current, broken machine to someone who can repair it.
  • New Wiki-pages:
  • Wiki security: For certain parts of the Wiki it would be convenient to have access control, so that only people from IRIDIA can access those pages. Anders will look into that.
  • E-pucks: All the available info on the e-pucks will be put on the Wiki.
  • Sbot/simulator interfaces: Rehan is going to create a repository for the common sbot interface and it is going to be implemented in TwoDee.
  • OMiss simulator: Rehan is going to create a Wiki page on checking out and compiling OMiss (the new simulator based on ODE).



Wednesday 19th October 2005 - Alexandre

Presentation / Discussion Topics

The main topic is optimal multi-foraging in swarm robotics.

  • I will present my latest developments and results
  • Go again through goals, motivations and bibliography (renewed).
  • Discussion about the context of this work, possible applications ?
  • I might present, if there is time left, my next experiment.

You can take a look at this very quick summary of my current research : http://iridia.ulb.ac.be/~acampo/various/alex_research.pdf

Administration Issues

  • Coffee machines
  • Format of these meetings
  • Standard programatic interface for simulators / real sbots.

People who will be absent

Rodi

Results

You can download a copy of Alex's presentation

  • The optimization group will have their own weekly presentations, but the will invite robotics people to participate, just like the swarm-bots presentations are open to optimization people. Seminars will continue as always and presentations might be given both at weekly meetings and at seminars.
  • Anders brings his coffee machine to test it. In case it pass the Italian test, we will buy one while the super-machine is being fixed.
  • Simulators and a standard interface for simulators:
    • Rehan and Alex are going to talk to Vito and make a plan
    • Rehan proposed Halva's interface, which he is going to send to Ciro and Anders for discussion on the next meeting

Wednesday 14th June 2006 - Alexandre

Agenda

  1. Presentation / Discussion
    • The presentation is about foraging. I present a mechanism that aims at selecting the closest resource for a group of robots.
  2. Robotics Group Admin

Reading material


People who will be absent

Roderich



Wednesday 21st June 2006 - Rehan

Agenda

  1. Presentation / Discussion
    • I will present the likely theme for my Phd: 'Swarm Control'. Research in Swarm robotics is currently focused on producing increasingly sophisticated autonomous behaviour in robotic swarms. However intelligent robot swarms become, humans will still for the forseeable future need some way of controlling these swarms for them to be useful 'tools'. The design of such control mechanisms is not obvious, since in distributed swarm systems there is no natural mapping from human goals to the robots' internal representation of the world.
  2. Robotics Group Admin
    • Sbot Common Repository - Rehan

Reading material


People who will be absent




Wednesday 28th June 2006 - Anders

Agenda

  1. Presentation / Discussion
    • Fault detection, isolation and tolerance for autonomous robots.
  2. Robotics Group Admin
    • Let's discuss Francesco request about new s-bots specifications and prepare a document for him (Marco)

Reading material


People who will be absent




Wednesday 5th July 2006 - Elio

Agenda

  1. Presentation / Discussion
    • I will try to present the results of the analysis of the mechanisms employed by the robots to solve the navigation task I described the previous time. My presentation will be very very informal. Please, do not expect too much since I'm particularly busy working on other issues
  2. Robotics Group Admin

Reading material


People who will be absent




Wednesday 15th November 2006 - Shervin

Agenda

  1. Presentation / Discussion
    • The presentation is about the experiment I'm currently setting up on simulation. The experiment compares the chains with a new controller I call forcefield. I will shortly describe my new controller, and give a detailed description of the experimental setup and the results I expect. My goal for the discussion is mainly to get some feedback and criticism concerning the experiment.
  2. Robotics Group Admin

Reading material


People who will be absent

Wednesday 6th December 2006 - Alexandre

Agenda

  1. Presentation / Discussion
    • The presentation will describe my current work on the use of virtual pheromones for path selection. The predictions of a mathematical model will be compared with results from simulated experiments. I will show that the group of robots is able to display a collective choice in favour of the shortest path, and that it can possibly balance the exploitation of these paths.
  2. Robotics Group Admin

Reading material

People who will be absent



Wednesday 10th January 2006 - Rehan

Agenda

I will be presenting the recent set of experiments I have done on functional self assembly and adaptive hill rotation.

  1. Presentation / Discussion
  2. Robotics Group Admin

Reading material

People who will be absent



Wednesday 17th January 2006 - Anders

Agenda

  1. Presentation / Discussion
    • Fault detection for automonous mobile robots.
  2. Robotics Group Admin

Reading material

People who will be absent




Wednesday 31st January 2006 - Christos

Agenda

  1. Presentation / Discussion
    • I will present my work in progress concerning :

a)The Evolution of Self-Assembly: Artificial Evolution can be used as an automated process that can identify the conditions for achieving self-assembly in autonomous robots. In this work we follow an approach where the experimenter does not predefine all necessary conditions,including motion, gripping, seeding. On the contrary, we want the ANN to control all these events. Particular importance will be given to the emergence of role in the robots. "Who grips whom" in a homogeneous robot group? b)The Evolution of Communication in the case of collective categorization through different history or experience. This work is an extension of an idea I developed on trying to evolve communication in groups of robots with different sensory apparatus (morphologically heterogeneous agents).

  1. Robotics Group Admin

Reading material

People who will be absent Rehan



Wednesday 14th February 2007 - ECAgents project DVD projection

Agenda

  1. Presentation / Discussion
    • The IST-FET project ECAgents (www.ecagents.org) is a project on which Elio and Christos are working. The project is coordinated by CNR in Rome (Stefano Nolfi) and participants include Luc Steel's SONY Lab in Paris, Jean-Louis Deneubourg in ULB, Dario Floreano at EPFL-Lausanne among others. Recently a DVD was made introducing the project, its goals and some of its results and this DVD will be projected so that people in IRIDIA know more details about ECAgents.
  2. Robotics Group Admin

Reading material

People who will be absent



Wednesday 7th March 2006 - Carlo

Agenda

I will present some papers I've happened to read and find inspiring in the last month.

  1. Presentation / Discussion
    • An introduction to Spiking Networks: a biological background and some applications to Robotics.
    • An interesting application for Collective Robotics: the Capture the Flag game.
  2. Robotics Group Admin

Reading material

Gerstner, Kreiter, Markram, Herz, Neural Codes: Firing Rates and Beyond, Proc. Natl. Acad. Sci. 1997
Abbot, Nelson, Synaptic Plasticity: Taming the Beast, Nature 2000
Floreano, Mattiussi, Evolution of Spiking Neural Controllers for Autonomous Vision-Based Robots, Springer 2001
Di Paolo, Spike-timing Dependent Plasticity for Evolved Robots, Adaptive Behavior 2002
Di Paolo, Evolving Spike-Timing-Dependent Plasticity for Single Trial Learning in Robots, Phil. Trans. R. Soc. Lond. 2003
Soula, Alwan, Beslon, Learning at the Edge of Chaos: Temporal Coupling of Spiking Neurons Controller for Autonomous Robotic, AAAI 2005
Nelson, Grant, Henderson, Evolution of Neural Controllers for Competitive Game Playing with Teams of Mobile Robots, Elsevier 2004

You can download the presentation at http://iridia.ulb.ac.be/~cpinciroli/extra/RoboticsMeeting0703.pdf

People who will be absent



Wednesday 14th March 2006 - Xavier

Agenda

  1. Presentation / Discussion
  2. Robotics Group Admin

Reading material

People who will be absent



Wednesday 28th March 2006 Post doc #2

Agenda

  1. Presentation / Discussion
  2. Robotics Group Admin

Reading material

People who will be absent