A case-based approach for coordinated action selection in robot soccer

Designing coordinated robot behaviors in uncertain, dynamic, real-time, adversarial environments, such as in robot soccer, is very challenging. In this work we present a case-based reasoning approach for cooperative action selection, which relies on the storage, retrieval, and adaptation of example cases. We focus on cases of coordinated attacking passes between robots in the presence of the defending opponent robots. We present the case representation explicitly distinguishing between controllable and uncontrollable indexing features, corresponding to the positions of the team members and opponent robots, respectively. We use the symmetric properties of the domain to automatically augment the case library. We introduce a retrieval technique that weights the similarity of a situation in terms of the continuous ball positional features, the uncontrollable features, and the cost of moving the robots from the current situation to match the case controllable features. The case adaptation includes a best match between the positions of the robots in the past case and in the new situation. The robots are assigned an adapted position to which they move to maximize the match to the retrieved case. Case retrieval and reuse are achieved within the distributed team of robots through communication and sharing of own internal states and actions. We evaluate our approach, both in simulation and with real robots, in laboratory scenarios with two attacking robots versus two defending robots as well as versus a defender and a goalie. We show that we achieve the desired coordinated passing behavior, and also outperform a reactive action selection approach.     

Gravity and solidity in four great ape species (Gorilla gorilla, Pongo pygmaeus, Pan troglodytes, Pan paniscus): Vertical and horizontal variations of the table task.

Three experiments modeled after infant studies were run on four great ape species (Gorilla gorilla, Pongo pygmaeus, Pan troglodytes, Pan paniscus) to investigate their reasoning about solidity and gravity constraints. The aims were: (a) to find out if great apes are subject to gravity biased search or display sensitivity for object solidity, (b) to check for species differences, and (c) to assess if a gravity hypothesis or more parsimonious explanations best account for failures observed. Results indicate that great apes, unlike monkeys, show no reliable gravity bias, that ape species slightly differ in terms of their performance, and that the errors made are best explained by a gravity account. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Effluent content from albedo degradation and kinetics at different temperatures in the enzymatic peeling of grapefruits

This work studies the variation in the effluent content from the reactor in enzymatic peeling of grapefruits, to determine the loss of soluble solids, reducing sugars and galacturonic acid, the main component from the albedo of the skin to model the solid–liquid transfer in this process. Experiments to study the efficiency of the enzymatic reaction include the effect of the temperature on the kinetics of albedo degradation to find an equation relating both variables and also the activation energy of this process, its value being 36.9 kJ/mol.     

An automatic and general least-squares projection procedure for sweep meshing

In this paper, we present a new and automatic node projection algorithm to generate hexahedral meshes in extrusion geometries. It is designed to preserve the shape of the cap surfaces in the inner layers of a sweeping mesh. The algorithm is based on least-squares approximation of affine mappings. We report that the functionals that have been traditionally used to compute the affine mapping generate four undesired effects on the inner layers of nodes. Namely, the flattening, skewing, offset scaling, and flipping of the inner layer meshes. To overcome these drawbacks we first analyze several properties of a new functional that depends on two vector parameters. Second, we introduce the concept of the pseudo-area and pseudo-normal vectors defined by a loop of nodes. In addition, we prove several geometrical properties of these vectors. Third, based on the properties of the new functional and on the definition of the pseudo-normal vector, we detail a new projection algorithm that automatically selects the functional vector parameters. The aim of this paper is to provide the implementation details to developers, although we also present the background of the algorithm. Finally, several mesh examples are discussed to assess the properties of the proposed algorithm.     

Architectures for Extreme-Scale Computing

Extreme-scale computers promise orders-of-magnitude improvement in performance over current high-end machines for the same machine power consumption and physical footprint. They also bring some important architectural challenges.     

Crosslinking of mixtures of diglycidylether of bisphenol‐A with 1,6‐dioxaspiro[4.4] nonan‐2,7‐dione initiated by tertiary amines: III. Effect of hydroxyl groups on network formation

BACKGROUND: Blends of epoxy resin oligomers, diglycidylether of bisphenol‐A (DGEBA), and a bislactone, 1,6‐dioxaspiro[4.4]nonan‐2,7‐dione (s(γBL)), were anionically copolymerized using two tertiary amines as anionic initiators. Their curing rheology and gelation behaviour were studied to provide a more comprehensive knowledge of the curing of these previously studied systems. RESULTS: The activation energy for gelation was found to be similar to that previously measured using differential scanning calorimetry and appeared to increase in the presence of the bislactone. The reaction rate during copolymerization of DGEBA with s(γBL) was slower than DGEBA homopolymerization alone because the alkoxide attack on the epoxide is faster than the reaction of the carboxylate ion and the epoxy group. The effect of the initiator type on the gel conversion was small and was presumably due to differences in the kinetic chain length caused by amine regeneration from the quaternary amine. For the same initiator and at a constant ratio of DGEBA/s(γBL), an increase in the hydroxyl concentration of the DGEBA oligomer raised the gel conversion. For a DGEBA oligomer with low hydroxyl levels, an increase in the concentration of s(γBL) increased the gel conversion; however, for a DGEBA oligomer with high hydroxyl levels, increasing s(γBL) concentration decreased the gel conversion. CONCLUSION: These results are interpreted in terms of the effect of initiation rate and chain transfer rate on the kinetic chain length. The glass transition temperature of the gel was found to be controlled by the fraction of the aliphatic s(γBL) and the amount of plasticizing sol in the matrix. Copyright © 2009 Society of Chemical Industry     

Convergent optical correlator aligment based on frequency filtering

The convergent correlator is widely used but it presents the drawback of the alignment requirements: fine focusing of the input scene Fourier transform on the filter plane, filter centering, scaling the scene Fourier transform to match the filter size, and azimuth matching of the filter with the input scene. We propose a set of tests to obtain a precise alignment of the convergent correlator. These methods are based on frequency filtering properties and they are applicable either for amplitude input or for phase-encoded input. The tests we present allow us to fulfill all the alignment requirements. The theory on which these tests are based is explained. The experimental results obtained during the alignment procedure are presented. We show some additional verifications of the correct alignment of the convergent correlator.     

Slicing XML Documents

Program slicing is a well-known technique to extract the program statements that (potentially) affect the values computed at some point of interest. In this work, we introduce a novel slicing method for XML documents. Essentially, given an XML document (which is valid w.r.t. some DTD), we produce a new XML document (a slice) that contains the relevant information in the original XML document according to some criterion. Furthermore, we also output a new DTD such that the computed slice is valid w.r.t. this DTD. A prototype implementation of the XML slicer has been undertaken.     

A Matrix-Based Approach to the Image Moment Problem

An image can be seen as an element of a vector space and hence it can be expressed in as a linear combination of the elements of any non necessarily orthogonal basis of this space. After giving a matrix formulation of this well-known fact, this paper presents a reconstruction method of an image from its moments that sheds new light on this inverse problem. Two main contributions are presented: (a) the results using the standard approach based on the least squares approximation of the result using orthogonal polynomials can also be obtained using matrix pseudoinverses, which implies higher control on the numerical stability of the problem; and (b) it is possible to use basis functions in the reconstruction different from orthogonal polynomials, such as Fourier or Haar basis, allowing to introduce constraints relative to the bandwidth or the spatial resolution on the image to be reconstructed. Judit Martònez received the B.Sc. degree in 1993 and the PhD degree (with honors) in 1998, both in telecommunications engineering from the Technical University of Catalonia. She developed her research at the Institut de Robútica i Informütica Industrial of the Spanish High Council for Scientific Research. In 1999 she joined the Computer Vision Center, a R&D center founded by the Autonomous University of Barcelona and the Autonomous Government of Catalonia. She has been principal researcher of several industrial and research projects related to computer vision technologies. Her research interests include industrial applications of machine vision, efficient algorithms for low-level image processing, multiresolution mathematical models, statistical clustering, pattern classification and inverse problems. Josep M. Porta received the Engineer Degree in Computer Science in 1994 and the Ph.D. in Artificial Intelligence in 2001, both from the Technical University of Catalonia (UPC). After that, he joined the IAS group of the University of Amsterdam and currently, he holds a post-doc position at the Institut de Robútica i Informütica Industrial (CSIC-UPC) in Barcelona. He carried research in legged robots, machine learning, vision-based methods for autonomous robot localization, and computational kinematics. Federico Thomas is Research Professor at the Spanish Scientific Research Council (CSIC) and director of the Institut de Robútica i Informütica Industrial (CSIC-UPC), Barcelona, Spain. He received the telecommunications engineering degree in 1984, and the Ph.D. degree (with honors) in computer science in 1988, both from the Technical University of Catalonia (UPC). In 1991, he won a NATO postdoctoral scholarship at the University of Massachusetts with the late Prof. Robin Popplestone. In 1999, he was visiting professor, sponsored by the Autonomous Government of Catalonia, at the Oxford University Computing Laboratory with Prof. Stephen Cameron. He has been project leader of several national projects financed by the Spanish Committee for Science and Technology (CICYT), and by local companies such as ENHER, a power generation company now part of ENDESA. His current research interests are in Geometry and Kinematics with applications to Robotics, Computer Graphics and Computer Vision. Prof. Thomas is an Associate Editor of the IEEE Transactions on Robotics.