DO ANDROIDS DREAM OF ELEC-tric sheep? Science fiction books, movies, and TV shows have long described humanoid robots interacting with-even passing forreal humans. But how close are we today to that sort of seamlessness between robots and humans?In exploring that question, we have developed a methodology to provide the basic mechanisms that support humanoid robot cognition. Involved are mechanisms for attention control and pattern categorization (see the "Attention" sidebar), and constructing attentional maps of the environmental context that underlies behavior. Our system architecture supports several types of sensors, and we have experimented with many facets of the eventual integrated platform for learning control, acquired representations, and visual behavior. [1][2][3][4] Our goal is to develop a robotic system capable of interacting with its environment and eventually with humans. We're also interested in the relationships between visual, haptic (touch-based), proprioceptive (simuli arising from within), and motor systems in humans. To this end, we constructed a robot, Magilla, consisting of a stereo head and two robotic arms (with attached graspers). To explore the potential for rich and varied interactions with the world, we connect our robot's cognitive organization and development to that of a human's-a methodology missing in the approaches the AI community usually employs. Moreover, by grounding our systems in the natural world (that humans share) and by providing flexible and redundant means of interacting within this domain, we postulate that our robot will develop cognitive structures more like those of humans and, as a consequence, will interact with humans more constructively. Learning from childrenAgain, one of our main objectives in constructing an anthropomorphic torso is to study the relationship between visual and haptic sensory systems and their development in humans.The development of robot programs is an incremental search for strategies that exploit the intrinsic dynamics of the robot-world interaction. We interpret intrinsic dynamics fairly broadly as any kinematic, dynamic, perceptual, or motor synergy that produces characteristic and invariant temporal sequences of observable state variables.Humanoid robots are simply too complex to use traditional approaches from robotics and computer vision. The range of interaction possible and the required kinds of perceptual distinction necessary challenge commonly used methodologies for control and programming. Consequently, we have adopted an incremental and automatic approach to programming, modeled after the sensorimotor development of human children in the first two years of life. Genetically mediated maturational mecha- HUMANOID ROBOTS PROMISE TO LEAD US TOWARD MORE EFFECTIVE AND INFORMATIVE INTERACTIONS IEEE INTELLIGENT SYSTEMSnisms focus the infant on simple problems first and subsequently enrich these policies by including additional motor and perceptual systems. 5 Infants are constantly learning about the capabilities of their motor sy...

Activities of glutathione S-transferases (GST) and cholinesterase (ChE) from Paracentrotus lividus were investigated as possible biomarkers of environmental contamination in the coastal zone. In the first phase of the study, the activity of both enzymes was determined in various tissues in order to select the most appropriate ones to be used in the following assays. In the second phase, the ChEs present in ambulacra were characterized using different substrates and selective inhibitors. In the next phase, laboratory bioassays were performed with dilutions of water-accommodated fraction of #4 fuel-oil (WAF) and benzo[a]pyrene (BaP) to determine the response of those enzymes to these pollutants and, finally, the activity of both enzymes was determined during a year in indigenous specimens from six sites on the Northwest coast of Portugal, with different pollution levels, to determine basal values and seasonal variations of ChE and GST activities. Among the several tissues tested, ambulacra and the anterior portion of the intestine were selected for ChE and GST assays, respectively. The determination of ChE in ambulacra tissue may be performed in a non-destructive way. Ambulacra ChE hydrolysed acetylthiocholine preferentially to propionylthiocholine and butyrylthiocholine and, inhibition by excess of substrate was observed. Enzymatic activity was almost fully inhibited by eserine sulfate (>98%) at concentrations equal or higher than 6.25 microM. Sensitivity to both BW284C51 (reaching 98% at 200 microM) and iso-OMPA (73% at 8 mM) was found. In laboratory bioassays, GSTs activity was inhibited by WAF and induced by BaP, whereas ChE activity was not affected by any of these environmental contaminants. Seasonal variations in enzymatic activities were found. For example, in a reference site, ChE values changed from 52.2 +/- 9.3 U mg(-1) protein in autumn to 71.8 +/- 13.3 U mg(-1) protein in spring, while GST activity changed from 129.9 +/- 29.8 U mg(-1) protein in winter to 279.0 +/- 48.0 U mg(-1) protein in autumn. Sea-urchins from reference sites presented significantly higher values of both ChE and GST than animals from contaminated sites in all seasons. In conclusion, the results of this study indicate that (i) ambulacra and the anterior portion of intestine are the most suitable tissues to measure ChE and GST activities, respectively; (ii) only one form of ChE seems to be present in ambulacra, showing properties of both typical acetylcholinesterase (AChE) and pseudocholinesterase (PChE); (iii)P. lividus GST is sensitive to both WAF and BaP even after acute exposures while ChE is not, and (iv) in spite of the significant seasonal variations observed in both enzymes in the field, P. lividus ChE and GST were capable of discriminate sites with different contamination levels and, thus, they are suitable for use as biomarkers in biomonitoring studies in the coastal zone.

With a view to using the cholinesterase (ChE) activity from guppy (Poecilia reticulata) muscle as a biomarker, the objectives of this work were: (i) to characterize the soluble cholinesterases present in muscle homogenate using different substrates and specific inhibitors, (ii) to determine the normal range of activity in non-exposed individuals and (iii) to investigate the in vitro effects of two common environmental contaminants, copper sulphate and dodecylbenzene sulphonic acid sodium salt (DBS) on ChE activity. The rate of substrate hydrolysis of P. reticulata ChE decreased in the order acetylthiocholine, propionylthiocholine and butyrylthiocholine. Inhibition by excess of substrate was observed at concentrations higher than 1.28 mM. Furthermore, eserine sulphate and 1,5-bis(4-allyldimethylammoniumphenyl)-pentan-3-one (BW284C51) significantly inhibited the enzyme activity at low concentrations (mM range) and N,N'-diisopropylphosphorodiamic acid (iso-OMPA) had no significant effect up to 8 mM. These findings suggest that the enzyme measured in this study is acetylcholinesterase. The activity determined in non-exposed fish was 145.1 ± 44.7 SD U mg(-1) protein. The common environmental contaminants copper and DBS significantly inhibited P. reticulata ChE at concentrations that can be ecologically relevant.

No abstract

Xenobiotic metabolising enzymes, namely the cytochrome P450 monooxygenase system and glutathione S-transferase, as well as antioxidant enzymes were determined in the liver of 3 deepsea fish species (tepidion lepidion, Bathypterois mediterraneusand Coryphaenoides guenthen). These organisms were collected from 3 stations along the NW Mediterranean, at depths ranging from 1500 to 1800 m. The enzymatlc activities detected in deep-sea fish indicated the ability of these organisms to cope with pollutants and oxidative stress. Differences among the studied species may be related to habitat and diet, viz. the highest monooxygenase and antioxidant enzyme activities were observed in L. lepidion, a typical middle slope species, whereas the lowest activities were recorded for B. mediterraneus, an organism adapted to live at greater depths. Tissue concentration of persistent organochlorinated compounds, like polychlorinated biphenyls, DDTs and hexachlorobenzene, were in the lower range of those reported for coastal fish, and significant differences among sampling stations were not observed. Overall, the obtained results contribute to the knowledge of the NW Mediterranean deepsea basin and reflect the chronic contamination of the area in terms of organochlorinated compounds, rather than the influence of coastal discharges.