LIFELIKE Computing Systems
9th Edition in the Evolution of the Workshop Series of Autonomously Learning and Optimizing Systems (SAOS)
++ Submission deadline extended by two weeks until June 6, 2021 ! ++
Agenda of the workshop day July 19, 2021 (all times are CEST)
Openingchaired by Anthony Stein, University of Hohenheim, DE
10:00 - 10:05 Welcome Note of the LIFELIKE organizers
10:05 - 10:30 "Lifelike Computing Systems" (paper pdf)
presented by Jean Botev, University of Luxembourg, LU
10:30 - 11:30 LIFELIKE Keynote "Sharks, Zombies and Volleyball - (Super)Natural Computation Gone Wild"
by Dr. Felipe Campelo, Aston University, UK
Contributed Papers Ichaired by Anthony Stein, University of Hohenheim, DE
11.30 to 11.55 "Analysing Metaheuristic Components" (paper pdf)
presented by Helena Stegherr, University of Augsburg, DE
++ 12.00 to 12.30 Lunch break ++
Contributed Papers IIchaired by Sven Tomforde, Kiel University, DE
12.30 to 12.55 "What I Want in a (Computational) Partner" (paper pdf)
presented by Christopher Landauer, Topcy House Consulting, CA
12.55 to 13.20 "Learning Classifier Systems for Self-Explaining Socio-Technical-Systems" (paper pdf)
presented by Michael Heider, University of Augsburg, DE
13.20 to 13.45 "A Concept for Self-Explanation of Macro-Level Behaviour in Lifelike Computing Systems" (paper pdf)
presented by Sven Tomforde, Kiel University, DE
Closing chaired by Peter Lewis, Ontario Tech University, CA
13.45 to 14.25 Discussion
14.25 to 14.30 Summary and Good bye
"Sharks, Zombies and Volleyball - (Super)Natural Computation Gone Wild"
by Dr. Felipe Campelo, Aston University, UK
Abstract: The field of meta-heuristic search algorithms has a long history of finding inspiration in natural systems. Starting from classics such as Genetic Algorithms and Ant Colony Optimization, we have recently witnessed an explosion of natural (and sometimes supernatural) methods in the literature - from Birds and Barnacles to Zombies and Reincarnation. While metaphors can often be powerful inspiration tools, it can be argued that the emergence of hundreds of barely discernible algorithmic variants under different labels and nomenclatures has done little to advance science, be it in terms of understanding and simulating biological systems, or of contributing generalisable knowledge or design principles for global optimisation approaches. In this talk we'll visit the Evolutionary Computation Bestiary and use it to discuss some of the possible causes and consequences of this trend, as well as some efforts aimed at moving the area of metaheuristics towards a better balance between inspiration and scientific soundness.
Short Bio: Felipe Campelo is a Senior Lecturer in Computer Science at Aston University. He received his BSc in Electrical Engineering from UFMG/Brazil in 2003, and his MSc (Information Science and Technology) and PhD (Systems Science and Informatics) from Hokkaido University/Japan in 2006 and 2009, respectively. His research focuses on the development of integrated solution pipelines for prescriptive data analytics, seamlessly connecting data mining, multi-objective optimisation and decision support systems for the solution of a variety of problems in science and engineering, from electromagnetic devices and power distribution systems to bioinformatics. He is also involved with the development of methodologically and statistically sound protocols for the experimental comparison of optimisation algorithms. Felipe is a member of the IEEE Computational Intelligence Society, the ACM Special Interest Group on Genetic and Evolutionary Computation, and the Foundation for Open Access Statistics.
Aims and Scope of LIFELIKE
Complexity in Information and Communication Technology (ICT) is still increasing, driven by the growing number of devices with vast amounts of computational resources. These systems are also increasingly interwoven into the very fabric of society, playing a role in how we connect together and socialize, how we move, work, and do business, and considering the role of technology in the spread of information, even what we know. As a result, the control of these systems is polycentric and necessarily complex and adaptive. Approaches for control and governance go way beyond traditional notions of central administration, which we note is often simply impossible for human operators.
Based on these insights, a growing movement considers the necessity of capabilities that allow these systems to successfully act and survive in such complex, real environments -- they are supposed to be `lifelike'.
Inspired by organic systems, our future socio-technical and cyber-physical systems also need to exhibit such characterizing self-x properties. Research initiatives such as Organic Computing, Autonomic Computing as well as Self-aware Computing all share the common goal of understanding and engineering technical systems capable of dealing with uncertainty due to continually changing and highly dynamic environments.
In order to achieve the desired degree of system robustness and flexibility, the envisaged computing systems are:
Increasingly decentralized into large self-organizing collectives of (semi-)autonomous agents.
Equipped with sensors and actuators to perceive and modify their productive environment.
Deployed with machine learning, planning and optimization algorithms from the broad domain of Artificial Intelligence (AI) to render these (sub-)systems autonomously self-learning.
In light of this year’s ALIFE theme, we are also strongly convinced that beyond making systems `intelligent’ through AI technology, the Artificial Life community can deeply contribute to further advance the field of building viable future computing systems — which we call Lifelike Computing Systems.
Characterized by lifelike, or self-x properties such as being highly distributed and thereby acting self-motivated, self-organizing, self-adaptive, self-improving, self-healing, etc., Lifelike Computing systems foster a paradigm shift regarding their design and deployment. As a result, the vision of Organic and Autonomic Computing manifests itself — traditional design-time decisions are moved to the productive runtime and, thus, the systems themselves take over control. Although this would dramatically increase the degree of system autonomy, it also satisfies the conditions for emergence occurring. This aspect, however, should be envisioned as a double-edged sword, since emergent effects can be beneficial but also detrimental; at least for our technical computing systems. In any case, technical systems must comply with necessary safety boundaries apparent in nearly any real-world application with humans involved. Therefore, we deem Lifelike Computing Systems as urgently required to be self-explanatory!
This workshop is intended to provide a forum for discussing the implications and new insights from adopting Artificial Life principles to technical computing systems acting in real-world environments. Additionally, we explicitly emphasize the aspects of interpretability and explainability of the involved algorithms in order to provide a basis for system transparency already at the core of its mechanisms. Besides this self-explanatory property, further key ingredients to reach a specific level of intelligence are self-awareness and the resulting ongoing pursuit for continual self-improvement by means of learning and optimization. The resulting, particular tension between increasing system viability through adopting lifelike characteristics, while at the same time ensuring an appropriate degree of system explainability, validation and compliance to exploration boundaries, constitutes the main motivation and unique topic of this workshop.
Therefore, we solicit research and position papers which are expected to set their focus on at least one or else multiple self-x properties for realizing Lifelike Computing Systems, among others:
Self-organization, i.e., adoption of organic system principles concerning bottom-up evolution of system structures, holarchies, trusted communities or socio-technical design concepts.
Self-explainability, i.e., deriving new metrics for quantification, system validation, guaranteeing, understanding \& trust as well as proper ways for visualization via context-aware and transient interfaces.
Self-improvement, i.e., continual behavior optimization@runtime through mechanisms such as automated algorithm configuration \& selection or evolutionary intelligence as a mechanism to change.
Self-awareness, i.e., establishing autonomous learning behaviour in technical systems by means of techniques such as active learning, transfer learning, online concept drift and novelty detection, efficient reinforcement learning from feedback, or model self-reflection.
Submission deadline: May 24, 2021
Extended submission deadline: June 6, 2021
Decision notification: June 20, 2021
New decision notification: June 22, 2021 (due to submission deadline extension)
Papers must be written in English and are expected to report on innovative ideas and novel research results around the topic of LIFELIKE. Reported results and findings have to be integrated with the current state of the art and should provide details and metrics allowing for an assessment of practical as well as statistical significance. Contributions bringing in novel ideas and concepts from related fields such as Organic Computing, Autonomic Computing, Self-aware Computing, etc. are explicitly solicited, but authors are at the same time strongly encouraged to clearly state the relevance and relation to LIFELIKE's as well as ALIFE's main theme.
conform to the ALIFE submission instructions.
2 pages for extended abstracts on new ideas, notes, discussion points, or reporting on relevant work recently published elsewhere (e.g. in journals)
4 pages for position papers, raising intriguing standpoints/hypotheses or summarizing pursued / proposed research agendas
6 pages for research papers reporting on original results and novel insights underpinned by experimental or theoretical evidence.
be submitted via EasyChair.
Anthony SteinUniversity of Hohenheim (DE)
Sven TomfordeKiel University (DE)
Jean BotevUniversity of Luxembourg (LU)
Peter LewisOntario Tech University (CA)
Jacob Beal, BBN Technologies
Kirstie Bellman, Topcy House Consulting
Jean Botev, University of Luxembourg
Uwe Brinkschulte, University of Frankfurt
Ada Diaconescu, Telecom ParisTech
Lukas Esterle, Aston University
Jörg Hähner, University of Augsburg
Heiko Hamann, University of Lübeck
Christian Krupitzer, University of Hohenheim
Chris Landauer, Topcy House Consulting
Peter Lewis, Ontario Tech University
Erik Maehle, University of Lübeck
Christian Müller-Schloer, Leibniz University Hannover
Wolfgang Reif, University of Augsburg
Bernhard Sick, University of Kassel
Anthony Stein, University of Hohenheim
Sven Tomforde, Kiel University
Sebastian von Mammen, University of Würzburg
Torben Weis, University of Duisburg-Essen
Previous Editions of LIFELIKE/SAOS
LIFELIKE Computing Systems Workshop 2020: https://lifelikecs.organic-computing.de/2020-edition
SAOS Workshop 2019: https://www.organic-computing.de/saos19