Integrating Symbolic Discovery with Numerical Machine Learning for Autonomous Swarm Control

2008/10/1 - 2012/3/31

Japanese team
Einoshin Suzuki	Leader, Professor	Department of Informatics, ISEE, Kyushu University
Shin Ando	Assistant Professor	Department of Computer Science, Faculty of Engineering, Gunma University
Shigeru Takano	Assistant Professor	Department of Informatics, ISEE, Kyushu University
Nguyen Huy Thach	Doctor course student	Graduate School of Systems Life Sciences, Kyushu University
Asuki Kouno	Doctor course student	Graduate School of Systems Life Sciences, Kyushu University
Theerasak Thanomphongphan	Master course student	Gunma University
Daisuke Hoshino	Master course student	Gunma University
Shouta Nagafuchi	Bachelor course student	Faculty of Engineering, Kyushu University
Kouichiro Nishikawa	Bachelor course student	Faculty of Engineering, Kyushu University
Emi Matsumoto	Bachelor course student	Faculty of Engineering, Kyushu University

French team
Michèle Sebag	Leader, Research Director	TAO project, CNRS
Marc Schoenauer	Research Director	TAO project, INRIA
Nicolas Bredèche	Associate Professor	LRI, University of Paris 11

The objective of the project is to design efficient autonomous controllers for swarm robots, which realize high-level intelligence that none of them can realize individually by gathering to make a group (i.e. swarm) of many robots. Specifically, the project takes advantage of the worldwide expertise of the partners in the complementary domains of symbolic learning, data mining and background knowledge modeling on the Japanese side, and probabilistic learning and stochastic optimization on the French side.

Possible applications include the design of swarm strategies for collective problem resolution (e.g. patrolling in hostile landscapes, rescue and identification of priorities, monitoring natural and artificial environments).

Organized workshop: LEMIR 2009 International Workshop@Slovenia, LEMIR 2011 International Workshop@Canada

Selected outcomes:
1. Einoshin Suzuki, Hiroshi Hirai, Shigeru Takano: "Toward a Novel Design of Swarm Robots Based on the Dynamic Bayesian Network", Advances in Data Management". Springer Studies in Computational Intelligence, Vol. 223, pp. 299-310, Zbigniew W. Ras and Agnieszka Dardzinska (eds.), Springer-Verlag, November 2009.
2. Einoshin Suzuki, Shigeru Takano, Hiroshi Hirai: Toward Using Symbolic Discovery in Designing Controllers of Autonomous Swarm Robots, Proc. First International Workshop on LEarning and data Mining for Robotics (LEMIR), pp. 1-10, September 2009, Bled, Slovenia.
3. Einoshin Suzuki and Michèle Sebag: Preface, Proc. First International Workshop on LEarning and data Mining for Robotics (LEMIR), p. ii, September 2009, Bled, Slovenia.
4. Asuki Kouno, Shigeru Takano, and Einoshin Suzuki: "Constructing Low-cost Swarm Robots that March in Column Formation", Swarm Intelligence (ANTS), LNCS 6234, pp. 556-557, September 2010, Brussels.
5. Swagat Kumar, Nguyen Huy Thach, and Einoshin Suzuki: "Understanding the Behaviour of Reactive Robots in a Patrol Task by Analysing their Trajectories", Proc. 2010 IEEE/WIC/ACM International Conference on Intelligent Agent Technology (IAT), pp. 56-63, September 2010, Toronto.
6. Shin Ando, Theerasak Thanomphongphan, Daisuke Hoshino, Yoichi Seki, and Einoshin Suzuki: "ACE: Anomaly Clustering Ensemble for Multi-perspective Anomaly Detection", Proc. Eleventh SIAM International Conference on Data Mining (SDM), pp. 1-12, Phoenix/Mesa, Arizona, April 2011 (datasets).
7. Hiroshi Hirai, Shigeru Takano, Einoshin Suzuki: Simulating Swarm Robots for a Collision Avoidance Problem based on a Dynamic Bayesian Network, Proc. Tenth European Conference on Artificial Life (ECAL 2009), LNAI 5778, Springer-Verlag, pp.416-423, 2011 (in print).
8. Asuki Kouno, Jean-Marc Montanier, Shigeru Takano, Nicolas Bredeche, Marc Schoenauer, Michele Sebag, and Einoshin Suzuki: "On-board Evolutionary Algorithm and Off-line Rule Discovery for Column Formation in Swarm Robotics", Proc. 2011 IEEE/WIC/ACM International Conference on Intelligent Agent Technology (IAT 2011), pp. 220-227, Lyon, August 2011.
9. Emi Matsumoto, Michele Sebag, and Einoshin Suzuki: "Avoiding Humans with SVM - a Case of Small Autonomous Mobile Robot in an Office", Proc. 26th International Symposium on Computer and Information Sciences (ISCIS), pp. ?-?, London, September 2011 (accepted for publication).

News
2008/10/1	project launched
2008/11/14	Post-doc position available announcement for the Japanese team (pdf,ps)
2008/11/25	Post-doc position available announcement for the French team
2009/6/8	Calls for a Post-doc position are closed for both Japanese and French teams
2010/6/28	Web page updated
2011/6/2	A movie, several pictures, benchmark data has been posted to the Web site of LEMIR 2011 International Workshop@Canada as benchmark data for ML/DM researchers