Distributed Robust Multi-Robot Learning using Particle Swarm Optimization

The goal of this project is the automatic design of high-performing robust controllers for mobile robots using exclusively on-board resources. In our evaluative approach, a population-based, on-line machine-learning technique automatically shapes the robots’ behaviors by direct interaction with the real environment. This constitutes an expensive optimization problem as the time needed to evaluate candidate solutions is substantially larger than that required by the metaheuristic operators in the algorithm.
In order to accomplish our goal, we must address two research questions. Firstly, we need to determine how to optimally allocate evaluation time for fast and robust adaptation, as evaluation time is the most critical resource in the process. Secondly, we need to define effective information sharing and cooperation mechanisms in our distributed system, as excessive sharing could lead to early stagnation, while limited sharing may slow down the convergence of the algorithm.

Team and Collaborators

Sponsors and Research Period

National Center of Competence in Research Robotics (NCCR Robotics)  and SNSF grant

Videos

Video of 4 robots learning obstacle avoidance using distributed PSO

Publications

2016

Measurements of the Higgs boson production and decay rates and constraints on its couplings from a combined ATLAS and CMS analysis of the LHC pp collision data at $ \sqrt{s}=7 $ and 8 TeV

G. Aad; B. Abbott; J. Abdallah; O. Abdinov; B. Abeloos et al. 

Journal of High Energy Physics. 2016-08-05. Vol. 2016, num. 8, p. 45. DOI : 10.1007/JHEP08(2016)045.

Noise-Resistant Particle Swarm Optimization for the Learning of Robust Obstacle Avoidance Controllers using a Depth Camera

I. Navarro; E. Di Mario; A. Martinoli 

2016. 2016 IEEE Congress on Evolutionary Computation, Vancouver, BC, Canada, July 24-29, 2016. p. 685-692. DOI : 10.1109/CEC.2016.7743859.

Distributed Learning of Cooperative Robotic Behaviors using Particle Swarm Optimization

E. L. Di Mario; I. Navarro; A. Martinoli 

2016. International Symposium on Experimental Robotics, Marrakech, Morocco, June 15-18, 2014. p. 591–604. DOI : 10.1007/978-3-319-23778-7_39.

2015

Combined Measurement of the Higgs Boson Mass in $pp$ Collisions at $\sqrt{s}=7$ and 8 TeV with the ATLAS and CMS Experiments

G. Aad; B. Abbott; J. Abdallah; O. Abdinov; R. Aben et al. 

Physical Review Letters. 2015-05-14. Vol. 114, num. 19, p. 191803. DOI : 10.1103/PhysRevLett.114.191803.

Distributed Multi-Robot Learning using Particle Swarm Optimization

E. L. Di Mario / A. Martinoli (Dir.)  

Lausanne, EPFL, 2015. 

Distributed Particle Swarm Optimization – Particle Allocation and Neighborhood Topologies for the Learning of Cooperative Robotic Behaviors

I. Navarro; E. L. Di Mario; A. Martinoli 

2015. IEEE/RSJ International Conference on Intelligent Robots and Systems, Hamburg, Germany, September 28 – October 02, 2015. p. 2958-2965. DOI : 10.1109/IROS.2015.7353785.

Distributed vs. Centralized Particle Swarm Optimization for Learning Flocking Behaviors

I. Navarro; E. Di Mario; A. Martinoli 

2015. 13th European Conference on Artificial Life (ECAL 2015), York, United Kingdom, 20-24 July 2015. p. 302-309. DOI : 10.7551/978-0-262-33027-5-ch056.

A Distributed Noise-Resistant Particle Swarm Optimization Algorithm for High-Dimensional Multi-Robot Learning

E. L. Di Mario; I. Navarro; A. Martinoli 

2015. IEEE International Conference on Robotics and Automation, Seattle, Washington, USA, May 26-30, 2015. p. 5970-5976. DOI : 10.1109/ICRA.2015.7140036.

SwarmViz: An Open-Source Visualization Tool for Particle Swarm Optimization

G. Jornod; E. L. Di Mario; I. Navarro; A. Martinoli 

2015. IEEE Congress on Evolutionary Computation, Sendai, Japan, May 25-28, 2015. p. 179-186. DOI : 10.1109/CEC.2015.7256890.

Distributed Particle Swarm Optimization using Optimal Computing Budget Allocation for Multi-Robot Learning

E. L. Di Mario; I. Navarro; A. Martinoli 

2015. IEEE Congress on Evolutionary Computation, Sendai, Japan, May 25-28 2015. p. 566-572. DOI : 10.1109/CEC.2015.7256940.

2014

Analysis of Fitness Noise in Particle Swarm Optimization: From Robotic Learning to Benchmark Functions

E. L. Di Mario; I. Navarro; A. Martinoli 

2014. IEEE Congress on Evolutionary Computation, Beijing, China, July 6-11, 2014. p. 2785-2792. DOI : 10.1109/CEC.2014.6900514.

The Role of Environmental and Controller Complexity in the Distributed Optimization of Multi-Robot Obstacle Avoidance

E. Di Mario; I. Navarro; A. Martinoli 

2014. IEEE International Conference on Robotics and Automation, Hong Kong, China, May 31 – June 7, 2014. p. 571-577. DOI : 10.1109/ICRA.2014.6906912.

Distributed Particle Swarm Optimization for limited-time adaptation with real robots

E. Di Mario; A. Martinoli 

Robotica. 2014. Vol. 32, num. 2, p. 193-208. DOI : 10.1017/S026357471300101X.

Distributed Particle Swarm Optimization for Limited Time Adaptation in Autonomous Robots

E. Di Mario; A. Martinoli 

2014. International Symposium on Distributed Autonomous Robotic Systems, Baltimore, Maryland, USA, November 8-11, 2012. p. 383-396. DOI : 10.1007/978-3-642-55146-8_27.

2013

The Effect of the Environment in the Synthesis of Robotic Controllers: A Case Study in Multi-Robot Obstacle Avoidance using Distributed Particle Swarm Optimization

E. Di Mario; I. Navarro; A. Martinoli 

2013. 12th European Conference on Artificial Life, Taormina, Italy, September 2-6, 2013. p. 561-568. DOI : 10.7551/978-0-262-31709-2-ch081.

A Comparison of PSO and Reinforcement Learning for Multi-Robot Obstacle Avoidance

E. Di Mario; Z. Talebpour; A. Martinoli 

2013. IEEE Congress on Evolutionary Computation, Cancún, México, June 20-23, 2013. p. 149-156. DOI : 10.1109/CEC.2013.6557565.

2011

A Trajectory-based Calibration Method for Stochastic Motion Models

E. Di Mario; G. Mermoud; M. Mastrangeli; A. Martinoli 

2011. IEEE/RSJ International Conference on Intelligent Robots and Systems. p. 4341-4347. DOI : 10.1109/IROS.2011.6094940.