EPS 2016 Workshop


We use the term Evolution of Physical Systems (EPS) to refer to evolutionary algorithms which occur entirely in real-world physical substrates rather than in simulation. The term encompasses both parallel Embodied Evolution (Watson et al., 2002), in which evolution is distributed across a population of robots, as well as Evolutionary Robotics (Floreano and Mondada, 1994) where evaluation is serialized on a single robot. Notable examples of EPS occur across a wide variety of systems, ranging from Robotics (Zykov et al., 2004)] to FPGAs (Thompson, 1996) to 3D printers (Rieffel and Sayles, 2010). Although EPS comes at a cost (the speed of the real world, unlike CPUs, does not follow Moore’s Law), by definition it avoids the “reality gap” imposed by simulation, and has produced novel and tangeable real-world results. Regardless of application or method, all implementations of EPS are bound by many of the same constraints and technical challenges.

The aim of this half-day single-track workshop is to bring together researchers who are currently involved in the Evolution of Physical Systems, as well as those interested in the technique, in order to share ideas and innovations. As the frontiers of artificial life move from the computer to the petri dish, the Evolution of Physical Systems offers to provide inroads into domains which are otherwise impossible to simulate.

The first four EPS Workshops at ALIFE13,  ECAL ’13, and ALIFE14 and ECAL15 were tremendous successes. We hope to continue the tradition at ALIFE ’16 in Cancun.

For this workshop, we are inviting participants and attendees to submit a draft paper for our upcoming special issue of Artificial Life Journal (MIT Press) devoted to the Evolution of Physical Systems, to be published November 2016.

How to Submit: E-mail the workshop chair (rieffelj@union.edu) with your name, contact info, and a PDF containing your either an extended abstract or complete draft of your special page submission.

Dates

  • Submission Deadline: 15April, 2015
  • Author Notification: 15 May, 2015

Organizing Committee:

John Rieffel, Union College, New York, USA
Jean-Baptiste Mouret, Universite Pierre et Marie Curie<br> Paris, France
 
Nicolas Bredeche, Universite Pierre et Marie Curie, Paris, France
 
Evert Haasdijk, University of Amsterdam Amsterdam, Netherlands

Call for Submissions:

Artificial Life Journal Special Issue on the Evolution of Physical Systems

We are inviting submissions to a forthcoming special issue of Artificial Life Journal devoted to the topic of the Evolution of Physical Systems (EPS).   This special issue emerges from a series of successful workshops held at the International Society for Artificial Life (ISAL) sponsored conferences ECAL and ALIFE.  We use the term EPS to refer to evolutionary algorithms which occur  in real-world physical substrates rather than in simulation. The term encompasses both parallel Embodied Evolution (Watson et al., 2002), in which evolution is distributed across a population of robots, as well as Evolutionary Robotics (Floreano and Mondada, 1994) where evaluation is serialized on a single robot. Notable examples of EPS occur across a wide variety of systems, ranging from Robotics (Zykov et al., 2004)] to FPGAs (Thompson, 1996) to 3D printers (Rieffel and Sayles, 2010).

While submissions describing reality-based and model-free work is encouraged, we would also be interested in submissions aimed at crossing the “Reality Gap” between simulation and reality.

Please direct all correspondence to rieffelj@union.edu

 

IMPORTANT DATES

  • Submission Deadline: April 15, 2016 May 15, 2016 EXTENDED!
  • Author Notification: July 1, 2016
  • Expected Publication: November 1, 2016

2015 Evolution of Physical Systems Workshop

Schedule and Agenda

 

14:00-14:05

Welcoming Remarks

John Rieffel, Union College


 

14:05-14:30

Roles of Model-Free Experiments in Evolutionary Robotics

Fumiya Iida and Luzius Brodbeck, Department of Engineering, University of Cambridge

The recent progress of robotics and IT technologies has enabled us to explore more flexible physically evolving systems. In our laboratory, for example, we have been taking advantage of affordable robotic manipulators and other automation technologies such as gluing machines and computer vision analysis, for the purpose of model-free evolution of physical robots [Brodbeck, et al. 2015].  Our platform is capable of handling cubic robotic modules, and gluing them in variations of morphological configurations such that constructed structures could exhibit diverse system-environment interactions such as dynamic crawling locomotion. The interactions can be also visually analyzed automatically, and the fitness evaluation can be used to improve morphological and controller designs in the real-world setup without any simulation models. Because most of the processes are automated, we are so far able to a large number of physical robots within a relatively short period of time (in the scale of hundreds of robots in a few weeks). Though the automation of robot construction and analyses of them is an interesting challenge by itself, we have been repeatedly questioned why simulated robot evolution is not sufficient. In this presentation, we would like to discuss a few distinctively important reasons why physical experiments of robot evolution are necessary, i.e. how the specifics of the real world drives evolution with interesting dynamics; how the real world implementation determines the level of abstraction; and how the details of real-world implementation can be compared to biological evolution in nature.


14:35-15:00

Evolving Computational Solutions in Matter

Julian Miller, University of York

Natural Evolution has been exploiting the physical properties of matter since life first appeared on earth.  Evolution-in-materio (EIM) attempts to follow this example by programming matter so that computational problems can be solved.  The beauty of this approach is that artificial evolution may be able to utilize
unknown physical effects to solve computational problems. This methodology is currently being undertaken in a European research project called NASCENCE: Nanoscale Engineering for Novel Computation using Evolution. We show how a variety of solutions to computational problems have been evolved using mixtures of carbon nanotubes and polymers at room temperature and also with gold nanoparticles at temperatures less than one Kelvin.


 

15:05-15:30:

Going from {0, 1} to [0, 1]

Eivind Samuelsen and Kyrre Glette Department of Informatics University of Oslo

When working with physical systems, one has to deal with variance in performance measurements. The real world is filled with noise, imperfections and chaotic behavior that introduce unwanted measurement error. It is common to measure performance over long time intervals or average a large number of measurements in order to filter out this variance. This makes experiments take a long time, and may prohibit some kinds of experiments altogether. We argue that research into smarter, more efficient use of measurement data may significantly reduce the time needed for each complete test. This would of course improve evaluation budgets of existing evolutionary algorithms on physical systems, and may also help realize applications such as rapid on-line gait learning and adaption.

At present, immense computing power is available to us, even when carried on-board freely moving robots or similar systems. We suggest using this computing power to apply statistical methods like resampling methods, regression and hypothesis testing. By collecting more data in the same time frame, and using statistics to analyze it more thoroughly, we aim to reduce the time needed to do individual measurements, while maintaining sufficient measurement accuracy.

For example, while measurements often need to be done over time in- tervals of a certain minimum length, say one control system period, they might not need to be disjoint intervals in order to supply some useful sta- tistical information. By sampling often, and then computing the perfor- mance for many feasible subintervals, i.e. resampling the data, we can obtain a large sample of data that can inform us about the variance of our measurements, and can give us better performance estimates.

Having a data sample associated with each measurement, not just a scalar performance value, also enables us to use statistical methods to bet- ter compare noisy measurements. This gives us probability estimates of one measurement outperforming another instead of a binary true/false answer. Although these estimates may be inaccurate, they are likely to be no less misleading than a scalar comparison. Such estimates can also be used by algorithms to decide whether more evaluation time is needed before making a decision, or if an evaluation can be stopped early.

When comparisons result in probabilities instead of binary choices, we need to adopt our algorithms accordingly. Evolutionary algorithms al- ready employ randomness extensively, so the changes needed should not be too big or too uncomfortable. Binary tournaments can be won by at random based on the probability of being best. Definitive rankings can be replaced with most-likely rankings. Fitness proportional mechanisms can be replaced with or supplemented by probability-of-being-best proportional equivalents.

It is our opinion that gathering and making better use of more data, and making algorithms aware of measurement uncertainties, can reduce the time needed to make measurements sufficiently accurate and improve on what sufficiently accurate is. Experimental verification of these ideas is needed to quantify the effects a more advanced statistical approach to individual measurements.

 


15:30-16:00 Tea Break


16:05-16:30

Evolving gaits for physical robots with generative encoding: using single-unit pattern generators

Danesh Tarapore, Antoine Cully, and J.-B Mouret, UPMC


 

16:35-17:00

Evolution and Analysis of Morphological Computation in a Physical Tensegrity Robot

John Rieffel, Union College


 

17:00

Panel Discussion and Wrap-Up

 

 

We use the term Evolution of Physical Systems (EPS) to refer to evolutionary algorithms which occur entirely in real-world physical substrates rather than in simulation. The term encompasses both parallel Embodied Evolution (Watson et al., 2002), in which evolution is distributed across a population of robots, as well as Evolutionary Robotics (Floreano and Mondada, 1994) where evaluation is serialized on a single robot. Notable examples of EPS occur across a wide variety of systems, ranging from Robotics (Zykov et al., 2004)] to FPGAs (Thompson, 1996) to 3D printers (Rieffel and Sayles, 2010). Although EPS comes at a cost (the speed of the real world, unlike CPUs, does not follow Moore’s Law), by definition it avoids the “reality gap” imposed by simulation, and has produced novel and tangeable real-world results. Regardless of application or method, all implementations of EPS are bound by many of the same constraints and technical challenges.

The aim of this half-day single-track workshop is to bring together researchers who are currently involved in the Evolution of Physical Systems, as well as those interested in the technique, in order to share ideas and innovations. As the frontiers of artificial life move from the computer to the petri dish, the Evolution of Physical Systems offers to provide inroads into domains which are otherwise impossible to simulate.

The first three EPS Workshops at ALIFE13,  ECAL ’13, and ALIFE14 were tremendous successes. We hope to continue the tradition at ECAL ’15. We will have time for ~6 speakers, as well as an extended panel session on the promises and pitfalls of the Evolution of Physical Systems. In addition, this year we will also feature a poster session.

We invite authors to submit a 1-page abstract describing research or a review/discussion related to the Evolution of Physical Systems.

Following the workshop, we will invite participants and attendees to submit to an upcoming special issue of Artificial Life Journal (MIT Press) devoted to the Evolution of Physical Systems.

How to Submit: E-mail the workshop chair (rieffelj@union.edu) with your name, contact info, and a PDF containing your 1 page submission.

Dates

  • Submission Deadline: 18 May, 2015
  • Author Notification: 25 May, 2015

Organizing Committee:

John Rieffel, Union College, New York, USA
Jean-Baptiste Mouret, Universite Pierre et Marie Curie, Paris, France
Nicolas Bredeche, Universite Pierre et Marie Curie, Paris, France
Evert Haasdijk, University of Amsterdam Amsterdam, Netherlands
Skip to toolbar