Catálogo de publicaciones - libros

Superstrings Installations (Torino 2005, Cetraro 2005, Milano 2005) – where Art and Science were directly conjugated – pictures and movies have been later collected into a DVD produced in a strictly limited edition and in few copies. These DVDs have been part of a Generative Art Installation presented in Milano in December 2006. A small robotic artifact has been programmed with very simple rules to trace on the DVD surfaces space “visual superstrings”, which changed depending both on the interaction with the public and the configuration of the DVDs on the floor. At the end, a unique artwork was generated. This paper discusses the ideas of A-Life Art, the related aesthetic concepts and how the use and programming of a robot has allowed to create A-Life Artistic object inspired by the Superstring Installation through semi-adaptive robotic behaviour.

In this paper we present an extension of the ShapeBugs distributed formation algorithm which enables 2D mobile agents to agree on a consensus coordinate system starting from no coordinate agreement. The participating agents require only local communication and local distance and motion information. Because this distance and motion information can in many cases be obtained implicitly through software motor approximation and attenuation or time-of-flight in communication, this approach has the potential to globally coordinate general mobile communicating agents without additional sensor requirements. The extended algorithm also remains robust against agent incapacitation and disorientation.

We develop a multi-level selection model in the framework of indirect reciprocity. Using two levels of selection, one at the individual level and another at the group level, we propose a competitive scenario among social norms, in which all individuals in each group undergo pairwise interactions, whereas all groups also engage in pairwise conflicts, modeled by different games. Norms evolve as a result of groups’ conflicts whereas evolution inside each group promotes the selection of best strategies for each ruling social norm. Different types of inter-group conflict and intensities of selection are considered. The proposed evolutionary model leads to the emergence of one of the recently obtained social norms, irrespective of the type of conflict between groups employed. We also compared the individual performance of the norm obtained in the evolutionary process with several other popular norms, showing that it performs better than any the other norms. This reputation assignment rule gives rise to a stern and unambiguous response to each individual behavior, where prompt forgiving coexists with implacable punishment.

Often the selfish and strong are believed to be favored by natural selection, even though cooperative interactions thrive at all levels of organization in living systems. Recent empirical data shows that networks representing the social interactions between people exhibit typically high average connectivity and associated single-to-broad-scale heterogeneity, a feature which precludes the emergence of cooperation in any static network. Here, we employ a model in which individuals are able to self-organize both their strategy and their social ties throughout evolution, based exclusively on their self-interest. The entangled evolution of individual strategy and network structure provides a key mechanism toward the sustainability of cooperation in social networks. The results show that simple topological dynamics reflecting the individual capacity for self-organization of social ties can produce realistic networks of high average connectivity with associated single-to-broad-scale heterogeneity, in which cooperation thrives.

We simulate the prisoner’s dilemma and hawk-dove games on a real social acquaintance network. Using a discrete analogue of replicator dynamics, we show that surprisingly high levels of cooperation can be achieved, contrary to what happens in unstructured mixing populations. Moreover, we empirically show that cooperation in this network is stable with respect to invasion by defectors.

In robotics, exogenous fault detection is the process through which one robot detects faults that occur in other, physically separate robots. In this paper, we study exogenous fault detection in a collective leader-follower task for autonomous robots. We record sensory inputs from the robots while they are operating normally and after simulated faults have been injected. Given that faults are simulated, we can correlate the flow of sensory inputs with the fault state of the robots. We use back-propagation neural networks to synthesize fault detection components. We show that the flow of sensory inputs is sufficient information for performing exogenous fault detection, that is, we show that the leader robot is capable of detecting faults in the follower robot. All results are based on experiments with real robots.

We identify two distinct themes in social science modelling. One, more specific, approach is that of social simulation which addresses how behaviour of many actors can lead to emergent effects. We argue that this approach, while useful as a tool in social science policy development, is fundamentally constrained due to the fact that its models are developed within the society they are supposed to model. Alternatively, the second theme looks to take a more holistic view by taking inspiration from systems sociology. This approach looks to build societies from the bottom up and may allow us to generate new perspectives in social theory.

In a social scenario, establishing whether a collaboration is required to achieve a certain goal is a complex problem that requires decision making capabilities and coordination among the members of the group. Depending on the environmental contingencies, solitary actions may result more efficient than collective ones and vice versa. In robotics, it may be difficult to estimate the utility of engaging in collaboration versus remaining solitary, especially if the robots have only limited knowledge about the environment. In this paper, we use artificial evolution to synthesise neural controllers that let a homogeneous group of robots decide when to switch from solitary to collective actions based on the information gathered through time. However, being in a social scenario, the decision taken by a robot can influence—and is influenced itself—by the status of the other robots that are taking their own decisions at the same time. We show that the simultaneous presence of robots trying to decide whether to engage in a collective action or not can lead to cooperation in the decision making process itself.

It is well known that certain environmental conditions, such as a spatially structured population, can promote the evolution of cooperative traits. However, such conditions are usually assumed to be externally imposed. In this paper, we present a model that allows the conditions that promote or hinder cooperation to arise adaptively via individual selection. Consequently, instead of selection simply favouring cooperation under imposed environmental conditions, in our model selection also operates on the conditions themselves via a niche construction process. Results are presented that show that the conditions that favour cooperation can evolve, even though those that favour selfish behaviour are also available and are initially selected for.

Pioneer approaches to Artificial Intelligence have traditionally neglected, in a chronological sequence, the agent body, the world where the agent is situated, and the other agents. With the advent of Collective Robotics approaches, important progresses were made toward embodying and situating the agents, together with the introduction of collective intelligence. However, the currently used models of social environments are still rather poor, jeopardizing the attempts of developing truly intelligent robot teams. In this paper, we propose a roadmap for a new approach to the design of multi-robot systems, mainly inspired by concepts from Institutional Economics, an alternative to mainstream neoclassical economic theory. Our approach intends to sophisticate the design of robot collectives by adding, to the currently popular emergentist view, the concepts of physically and socially bounded autonomy of cognitive agents, uncoupled interaction among them and deliberately set up coordination devices.