Category:PhDSupervision Andrea Giovanni Reina

Ph.D. Student:	Andreagiovanni Reina
Ph.D. started on:	1st May, 2011
Phone number (office):	+32-2-650 31 65
Phone number (mobile):	+32 487 403 588

Structure of the page.
Monthly report of activities with indication of percentage of time dedicated to each activity. The percentage is referred to the working time of one month (on average 22 days), that means that for example 10% correspond to 2.2 days, and 23% is one week (5 days). For simplicity I use 25% as one week, 50% as two weeks, and so on.

Sometimes the sum of all the activities exceed the 100%, it means that I worked much more than the regularly 8 hours per day.

May 2011

Project / Time:
5% Starting bureaucracy, setup, integration in the Lab.
50% Validation of the foot-bot's sensors.
25% E-SWARM website.
20% Swarm robotics literature.

Description:
Validation of the foot-bot's sensors. The goal was to model the precision (and the noise) of all the foot-bot's sensors. I was supervised by Carlo. We tried different solutions for the modelling the light sensors. All the attempts for having a reliable automatic method have failed. We abandoned the task.

E-SWARM website. Design, implementation, test and deployment of the E-SWARM website (http://www.e-swarm.org).

Swarm robotics literature. Reading papers and books about swarm robotics.

June 2011

Project / Time:
55% Proving that ARGoS simulates well.
25% Arena tracking system.
20% Swarm robotics literature.

Description:
Proving that ARGoS simulates well. Carlo needed to have an experiment where the simulated robot experiments coincide with the real robot experiments. Carlo needed that the proof of concept experiment contains heterogeneous swarms that collaborate. One of the past (unpublished) experiments of Carlo deals with the collaboration of heterogeneous swarms: A swarm of aerial robots that guide a swarm of ground robots. We decide to use this work for validating the simulator.
I start to collaborate with Carlo in order to perform real robot experiments and, in this way, validate the ARGoS simulator.

Arena tracking system. Learning and practicing with the HALCON software.
2 days of course at DataVision offices in Vlaardingen, Netherlands.

Swarm robotics literature. Reading papers and books about swarm robotics.

July 2011

Project / Time:
60% A ceiling robot guides a swarm of ground robots.
10% Arena tracking system.
15% Swarm robotics literature.
15% Distributed Path Planning.
1 day of holiday taken.

Description:
A ceiling robot guides a swarm of ground robots. We have proved (in June) that on real robot the scenario we are studying can be executed, but we do not have clear the experiments we want to do.
We try to define possible questions we want to answer with our experiments.
We start trying solutions, getting results and thinking on the results.
The main activities were:

• design experiments
  
• implement controllers
  
• run cluster experiments
  
• plot results (statistics and visualization)
  

Arena tracking system. Test of the tracking software. I and Alessandro created the tags to apply on the footbots and we started the first tracking experiments.

Swarm robotics literature. Reading papers and books about swarm robotics.

Distributed Path Planning. Getting ready for the VERY SHORT visiting period at IDSIA. I tried to have everything ready for the real robot experiments. But not enough time. :( Many things still to do.

August 2011

Project / Time:
100% Distributed Path Planning
25% Holiday (5 days taken)

Description:
Visiting IDSIA for 3 weeks (working day and night).
Created a COMPLETE REAL ROBOT DEMONSTRATOR.

1 week holiday.
TAROS Conference in Sheffield, UK. (best poster award)

September 2011

Project / Time:
120% AAMAS paper: A ceiling robot guides a swarm of ground robots (work with Carlo and Mauro).

Description:
I put all my effort in this project. (Unfortunately) Carlo was not in Belgium, due to the combination of the (already planned) conference IROS and holidays. Therefore, I was working mainly alone with the weekly supervision of Mauro.
I did my best to turn the work to a valid scientific contribution. I tried to give a scientific sense to the work. I tried to find an interesting message and to design experiments that give results which support the message.
Unfortunately, one month of intensive work was not enough: the message (and the scientific contribution) was still small for being accepted in a conference of high level like AAMAS.

During the month of September, I carried out a set of activities, among others I list the main ones:

• Problem (re)definition
  
• Design of the experiments
  
• Implementation of the robotic controllers
  
• Execution of large set of experiments on the cluster
  
• Result data analysis (statistics and plots)
  
• Documentation
  

October 2011

Project / Time:
80% AAMAS paper: A ceiling robot guides a swarm of ground robots (work with Carlo and Mauro).
15% Arena tracking system. + ANTS 2012
5% Distributed Path Planning

Description:
AAMAS paper. Working on the conclusion of the work with Carlo. The writing activity was mainly carried out by Carlo. I took care about all the rest.
Here a brief list of the main activities:

• Writing of the first draft of the paper.
  
• Real robot experiments
  
• Execution of large set of experiments on the cluster
  
• Result data analysis (statistics and plots)
  
• Real robot video

Tracking system. I and Alessandro evaluated the performance of the tracking system and designed the hardware requirements for the new arena.

ANTS 2012. Web sites updates (PC members, CFP, invited talks, ...).

November 2011

Project / Time:
40% Distributed Path Planning
10% Master student supervision + Various tasks @IRIDIA.
50% Holiday (8 days taken + National/ULB holidays)

Description:
Distributed Path Planning. Adding a new small functionality. Restructuring the code. Preparing the code for running on the cluster.

Master student supervision. Started to work with 2 Master students for the 2 projects of master I proposed: Web Community for Robotic Software and A web visualizer for ARGoS.

Various tasks @IRIDIA. Tracking system. ANTS 2012. Marco Dorigo's web pages.

December 2011

Project / Time:
50% Distributed Path Planning
20% FNRS Application
5% Various tasks @IRIDIA
25% Holiday (0 days taken + National/ULB holidays)

Description:
Distributed Path Planning. Restructuring the code. Preparing the code for running on the cluster.

FNRS Application. Starting to clarify the ideas. Selecting the topic. Focus the attention on what I want to study. Started collaboration with Vito and Alexander.

Various tasks @IRIDIA. Master Student's supervision. Tracking system. ANTS 2012. Marco Dorigo's web pages.

---

Holiday Summary 2011: 14 days taken (1 day July + 5 days August + 8 days November).
Adding the holidays of beginning of January (4 days), I reach 18 days.

---

January 2012

Project / Time:
40% FNRS Application
10% ANTS 2012
25% Holiday (4 days taken + National/ULB holidays)

Description:
FNRS Application. Focusing on the topic. Firsts drafts. Weekly feedback from Vito and Alexander.

ANTS 2012. Publicity and Online Conference System setup.

February 2012

Project / Time:
90% FNRS Application
10% ANTS 2012 + Master Students

Description:
FNRS Application. Writing project proposal. Daily feedbacks from Vito and Alexander. Completion of the application. Submission. Crossed fingers.

Future months plans:

March 2012

Project / Time:
100% Distributed Path Planning

Description:
Distributed Path Planning. Running experiments on the cluster. Analysis of the results (statistics and plots). First draft (at least table of content). Main supervisor: Gianni.

April 2012

Project / Time:
100% Distributed Path Planning

Description:
Distributed Path Planning. Additional experiments if needed. Writing journal article.

May 2012

Project / Time:
30% Distributed Path Planning
30% Rapport d'avancement des recherches
30% AMOS: Adaptive Morphogenesis in Swarm Robotics

Description:
Distributed Path Planning. Complete the journal article.

Rapport d'avancement des recherches. Preparation of the thesis. Defence.

AMOS: Adaptive Morphogenesis in Swarm Robotics. Starting the project on Adaptive Morphogenesis. Definition of the work plan. (Probably) Starting with: (i) review of opinion dynamics models, (ii) implementation of state of the art of morphogenesis process on real robots.