October 29-31, 2018. Rome, Italy
Consiglio Nazionale delle Ricerche
Piazzale Aldo Moro 7
00185, Rome, Italy.
Via San Martino della Battaglia 44
00185, Rome, Italy.
For accommodation we suggest to book directly at a hotel of your choice. There are many hotels in the area of the conference, which can satisfy any price requirements. There are also several listings on airbnb.com.
The conference is located close to the San Lorenzo neighbourhood, which offers a very large selection of restaurants, pizzerias (eat in and take away) and local street food. You will for sure find something to satisfy your appetite! Look at the map and choose your favourite place. Beware that not all restaurants are open for lunch, but the offer is still very large.
|9:00-10:00||Invited plenary talk
From Slime Moulds to Soccermatics: how flow and feedback create dynamic problem solving
David J. T. Sumpter, Uppsala University, Sweden
|10:30-12:10||Session 1: Oral presentations
|12:10-14:00||Free time for lunch|
|14:00-15:40||Session 2: Oral presentations
|15:40-16:00||Session 3: Preview highlights
|16:00-18:30||Poster session 1 + drinks and appetisers: Papers and previews presented in Sessions 1, 2, and 3|
|9:00-10:00||Invited plenary talk
The Power of Meta
Holger Hoos, Universiteit Leiden, The Netherlands
|10:30-12:10||Session 4: Oral presentations
|12:10-14:00||Free time for lunch|
|14:00-15:40||Session 5: Special Keynote Session
The Physics of Collectives: The Rome School
|15:40-16:00||Session 6: Preview highlights
|16:00-18:30||Poster session 2 + drinks and appetisers: Papers and previews presented in Sessions 4 and 6|
|9:00-10:00||Invited plenary talk
Multi-level Modeling for Swarm Robotics: Case Studies, Lessons, and Challenges
Alcherio Martinoli, EPFL, Switzerland
|10:30-12:10||Session 7: Oral presentations
|12:10-14:00||Free time for lunch|
|14:00-15:20||Session 8: Oral presentations
|15:20-15:40||Session 9: Preview highlights
|15:40-18:00||Poster session 3 + drinks and appetisers: Papers and previews presented in Sessions 7, 8 and 9|
|18:00-18:30||Award Ceremony and conclusion|
The ANTS2018 registration fee is 450 EUR.
The conference fee includes:
Coffee breaks and a conference dinner will be offered by the organizing committee.
To secure your participation to the conference, download and fill the registration form and follow the instructions therein.
Moulds to Soccermatics: how flow and feedback create
dynamic problem solving
Abstract: I will start the talk by discussing a model of current-reinforced random walks. This is a central tool in understanding network formation and problem solving by slime moulds and ants. My main innovation will be to stress the importance of reinforcing based on current, rather than on density, as is done in many ACO approaches. We show how this solves linear programming problems in an entirely decentralised way. I finish with a more light-hearted discussion about how my work on collective animal behaviour inspired the study of football from a mathematical perspective.
David Sumpter has worked on collective behaviour of
everything from slime moulds, through ants and honey
bees, fish and birds, as well as humans. His work
combines mathematical modelling of these ‘swarms’ with
experimental work on the detailed interactions of
individuals. He has written over 100 articles in leading
journals and wrote the book 'Collective Animal
Behaviour' summarising the field.
His most recent book 'Soccermatics' takes a new look at the world's most popular game, showing how mathematics works inside the game. The book is in seven languages, including Italian! David speaks regularly at book and science festivals, has given Google and TEDx talks, and his work often appears in the media, including BBC, NPR, and ABC. He has written for the Economist, FourFourTwo magazine, Science Daily, Scientific American, the Guardian, The Daily Telegraph and many other news media. David has published around 100 articles in leading scientific journals, including Science, Proceedings of the National Academy of Sciences, and the Royal Society journals. He has co-authored work with scientists from every continent of the world, apart from Antarctica. You can follow him on Twitter @Soccermatics.
David lives in Sweden with his wife and two children. He is professor of Applied Mathematics in Uppsala. In his spare time, he trains the his son's football team Upsala IF 2005.
Multi-level Modeling for Swarm
Robotics: Case Studies, Lessons, and
Abstract: Technological advances in communication, embedded computing, energy storage, sensors and actuators enable an increasingly higher number of potential applications for swarm robotics systems. Such systems and their related methods become competitive when the individual robotic nodes are severely constrained in their resources by cost, volume, or mass considerations imposed by the targeted application. Such constraints typically result in an increased stochasticity of the node behavior that has to be captured with appropriate methods in order to obtain a more predictable and controllable behavior at the collective system level. In this seminar, I will focus on one particular recipe that allowed us to achieve such result in specific scenarios and under given assumptions: multi-level modeling. I will describe our multi-level modeling framework and support the discussion by leveraging multiple case studies, starting from seminal ones in collision avoidance and collaborative manipulation and ending with recent ones in self-assembly. Despite the experimental scenarios related to these case studies are characterized by different environmental templates and capabilities of the individual robotic nodes in terms of computation, mobility, sensing, and actuation, I will show that the main multi-level modeling principles remain the same and enable further insights in the behavioral analysis and synthesis of the swarm robotic systems. Finally, I will conclude my seminar with some of the lessons we learned over more than twenty years of research in this area and extrapolate some hints for future research directions to overcome limitations of the current multi-level modeling methods.
Bio: Alcherio Martinoli has a M.Sc. in Electrical Engineering from the Swiss Federal Institute of Technology in Zurich (ETHZ), and a Ph.D. in Computer Science from the Swiss Federal Institute of Technology in Lausanne (EPFL). He is currently an Associate Professor at EPFL, leading the Distributed Intelligent Systems and Algorithms Laboratory and serving as director of the Doctoral Program in Robotics, Control, and Intelligent Systems. Before joining EPFL he carried out research activities at the Institute of Biomedical Engineering of the ETHZ, at the Institute of Industrial Automation of the Spanish Research Council in Madrid, Spain, and at the California Institute of Technology, Pasadena, U.S.A. His research interests focus on methods to design, control, model, and optimize distributed cyber-physical systems, including multi-robot systems, sensor and actuator networks, and intelligent vehicles. Among other contributions, Alcherio Martinoli has been a pioneer in swarm intelligence by proposing innovative model-based and data-driven methods (e.g., multi-level modeling, noise-resistant distributed PSO) for swarm robotic systems.
The Power of Meta
Abstract:Algorithms increasingly control our world and shape the way we interact with it. Up to now, most of these algorithms have been designed and implemented manually, but this is rapidly changing, as machine learning techniques are gaining traction. In this talk, I will argue that the transformational impact of that change lies in the move from algorithms to meta-algorithms: algorithms that operate upon algorithms. I will give examples illustrating the rise and success of meta-algorithmic techniques, and discuss the consequences of fully embracing their benefits. Then, I will turn to an intriguing question: What happens if we move up one further level of "meta", to algorithms that operate on meta-algorithms? To answer this question, I will trace several lines of work exploring this idea, including recent work in automated machine learning (AutoML), and sketch where it can lead us.
Bio: Holger H. Hoos is Professor
of Machine Learning at Universiteit Leiden (the
Netherlands) and Professor of Computer Science at the
University of British Columbia (Canada), where he also
holds an appointment as Faculty Associate at the Peter
Wall Institute for Advanced Studies. He is a Fellow of
the Association for the Advancement of Artificial
Intelligence (AAAI) and past president of the Canadian
Association for Artificial Intelligence / Association
pour l'intelligence artificielle au Canada (CAIAC).
Holger's research interests span artificial intelligence, empirical algorithmics, bioinformatics and computer music. He is known for his work on machine learning and optimisation methods for the automated design of high-performance algorithms and for his work on stochastic local search. Based on a broad view of machine learning, he has developed - and vigorously pursues - the paradigm of programming by optimisation (PbO); he is also one of the originators of the concept of automated machine learning (AutoML). Holger has a penchant for work at the boundaries between computing science and other disciplines, and much of his work is inspired by real-world applications.
Exploring the adjacent possible:
play, anticipation, surprise
Abstract: Novelties occur frequently in our individual daily lives. We meet new people, learn and use new words, listen to new songs, watch a new movie, adopt a new technology. Such new experiences sometimes happen by chance. Often they are triggered by earlier new experiences, thus providing an effective correlation between their appearances. Historically the notion of the new has always offered challenges to humankind. What is new often defies the natural tendency of humans to predict and control future events. Still, most of the decisions we take are based on our expectations about the future. From this perspective a deep understanding of the underlying mechanisms through which novelties emerge and humans anticipate their occurrence is key to progress in all sectors of human activities. The problem of anticipation, i.e., how to cope with the unexpected, is one of the open problems for Artificial Intelligent machines too. The common intuition that one new thing often leads to another is captured, mathematically, by the notion of "adjacent possible", i.e., the set of all those things (ideas, linguistic structures, concepts, molecules, genomes, technological artefacts, etc.) that are one step away from what actually exists, and hence can arise from incremental modifications and recombination of existing material. In this talk I'll present a mathematical framework, describing the expansion of the adjacent possible, whose predictions are borne out in several data sets drawn from social and technological systems. Finally I'll discuss how games could represent a extraordinary framework to experimentally investigate basic mechanisms at play whenever we learn, create and innovate. A better understanding of the space of possibilities and how we explore is key to deploy human imagination, face the societal challenges of our era and conceive a better future. What is the structure of the space of possibilities? How do humans explore it? How do machines explore it? These are some of the questions I'll try to address. And those questions are relevant in many areas, for instance, how do we take decisions, how do we anticipate the impact of specific choices, how do we learn and create, how do we conceive new (sustainable) solutions.
Bio: Full Professor of Physics of Complex Systems at Sapienza University of Rome and Faculty of the Complexity Science Hub in Vienna. He is presently Director of the SONY Computer Science Lab in Paris where he heads the team on "Innovation, Creativity and Artificial Intelligence". His scientific activity is focused on the statistical physics of complex systems and its interdisciplinary applications. He coordinated several projects at the EU level and he recently coordinated the KREYON project (www.kreyon.net) devoted to unfolding the dynamics of innovation and creativity. Vittorio has published over 180 papers in internationally refereed journals and conference proceedings and chaired several workshops and conferences.
Collective behavior in animal groups: a
Abstract: Many animal aggregations display collective patterns on the large scale, ultimately due to the interactions between the individuals in the group. Recent findings on flocks of birds and swarms of insects show that these groups exhibit strong mutual correlations and quick mechanisms of information propagation, signatures of the efficient collective response to external perturbations. Besides, they obey static and dynamic scaling laws suggesting that we can use a statistical physics approach to describe the large scale, and define novel `classes' of behavior. We will review our current understanding of collective animal behavior and discuss how a physics based perspective, from experiments to modelling, can help to define a unified description for these systems.
Bio: Andrea Cavagna received his
PhD in theoretical physics and statistical field
theory at Sapienza University in 1998, working on
spin-glasses under the supervision of Giorgio
Parisi. After spending four years as a postdoc in the
UK (Oxford and Manchester), he moved back to Rome,
where he joined the Institute for Complex Systems of
the National Research Council. After studying for
about a decade the statistical mechanics of disordered
systems, his research interests shifted in the last
ten years to problems in physical biology. Together with
Irene Giardina, he leads a lab for the study of Collective
Behaviour in Biological Systems (COBBS), whose aim is
to obtain 3D experimental data in the field and to
develop new theory directly from the data. The COBBS
lab has been the first to combine the production of
large-scale data (groups of up to 3000 individuals)
with a theoretical approach inspired by statistical
physics and field theory. The principal systems of
interest of COBBS have been bird flocks and insect
swarms, although new projects on the collective
properties of stem cells colonies and on the swarming
dynamics of malaria mosquitoes are being initiated. He
is the author of more than 70 articles, which received
about 6000 citations.
Irene Giardina received a Ph.D. degree in theoretical physics from the University of Rome La Sapienza in 1998. From 1999 to 2001 she worked as post-doctoral fellow at the University of Oxford and at the Laboratoire de Physique Theorique, CEA Saclay, where she studied a variety of problems in disordered and complex systems. In 2001 she was appointed research scientist at the Institute for Complex Systems, of the National Research Council in Rome. Together with Andrea Cavagna, she set up a new lab dedicated to apply methods from statistical physics to study collective behavior in animal groups and biological systems. From 2013 she is Associate Professor at the Department of Physics, Sapienza University of Rome.
Light driven bacteria: a million
microswimmers with remote control
Abstract: Proteorhodopsin is a light driven proton pump which uses photon energy to pump protons out of the inner membrane of bacteria. The resulting electrochemical gradient can drive the rotation of the flagellar motor so that, in a way, proteorhodopsin puts a "solar panel" on every cell, allowing to remotely control swimming speeds with light. These light powered bacteria can be employed as controllable biological propellers inside bio-hybrid micromachines. The synthetic components are 3D printed microstructures having a rotating unit that can capture individual cells into an array of microchambers designed so that each cell contributes maximally to the applied torque. Using a spatial light modulator, we can address individual motors with tuneable light intensities, control their individual speeds and also synchronize a set of micromotors to rotate in unison. When freely swimming in a dense suspension, these photokinetic bacteria provide a light controllable active fluid, whose density can be accurately shaped in space and time through structured light patterns. We show that a homogeneous sea of these swimming bacteria can be made to morph quickly between complex shapes and, when employing a feedback control strategy, can be used to display accurate and detailed reproductions of grayscale density images.
Bio: Roberto Di Leonardo is
Professor of Experimental Condensed Matter Physics at
the Physics Department of Sapienza University in
Rome. He is interested in the origin, the consequences
and the applications of motion at the micron scale,
from Brownian motion to cell motility. To study that,
his lab builds digital microscopes that integrate
optical and computer hardware and where light can be
used for imaging, manipulation and fabrication of
microsystems in 3D.
Di Leonardo received a PhD in Physics from the University of L’Aquila working on supercooled liquids and glass transition. He then moved to Rome to join the Center for Soft Matter Research of the National Institute for the Physics of Matter. In 2005 he moved to the University of Glasgow where he became interested in the use of light as a tool for manipulating matter on a micrometric scale. Beginning in 2009, he undertook the study of flagellar propulsion, focusing in particular on the possibility of exploiting self-propelling bacteria as a source of work in miniaturized devices. He is the author of more than 90 research papers on topics ranging from experimental optics to theoretical statistical mechanics. He is Fellow of the School for Advanced Studies Sapienza (SASS) and an ERC Starting grant recipient (2012).
ANTS 2018, continuing a tradition started with ANTS 2002, assigns a "Best paper award" consisting of a sculpture of an ant expressly created for the ANTS conference by the Italian sculptor Matteo Pugliese.
The best paper award has been graciously supported by LNCS Springer.
The winning paper was:
Quality-Sensitive Foraging by a Robot Swarm through Virtual Pheromone Trails
Anna Font Llenas, Mohamed S. Talamali, Xu Xu, James A.R. Marshall, and Andreagiovanni Reina
Congratulation also to the other two nominees:
Self-Adaptive Quantum Particle Swarm Optimization for Dynamic Environments by Gary Pamparà and Andries Engelbrecht.
An Analysis of the Intelligent Water Drops Algorithm by Christian Leonardo Camacho Villalón, Marco Dorigo and Thomas Stützle
Each oral presentation will last 20 minutes sharp (15
minutes presentation plus 5 minutes for
questions-and-answers). A beamer will be available. A
computer is available running Windows7, and PowerPoint and
Acrobat Reader are installed. You can use this computer for
your presentation, or you can bring your own laptop.
Paper presented orally will be also presented during the poster session on the same day of the presentation. Please refer to the on-line program for the details.
Posters should be of size A0 portrait. There is no standard template for the poster, every author can choose what best fits his work. Material to fix the poster on the stand will be available.
Papers that are not presented orally will be introduced
by the author in a 2 minute highlight, using a SINGLE
slide. This slide must be sent to us in advance. We will
preload this slide onto the computer and project it for
you during your presentation.
You should let us have the one slide you intend to use for your highlight presentation by October the 19th. It should be a single page in pdf format, landscape view, and without animations. Please send this file by email to firstname.lastname@example.org. The subject of the email should be
"ANTS 2018: Highlight LAST_NAME_OF_FIRST_AUTHOR"The poster session will start right after the highlight session.