In this paper, we aim to demonstrate the potential for wider-ranging capabilities and ease of transferability of our recently developed decision-making architecture for human-robot collaboration. To this end, a somewhat related but different application-specific example from the generic one used in its development is chosen, a toy car assembling task in which a participant works together with a robot to perform the assembly task. In a “Wizard of Oz” fashion, a comparison is made between the participant’s reactions to working with the robot being controlled either by our architecture or by a human “Wizard” who is hidden from view. With regard to the generalisability of the architecture, we also wish to investigate whether specific models trained on the observed human behaviour in a generic assembly task also transfer to this more complex task. Therefore, pre-trained interaction models from a prior generic pick-and-place task are used again in this new application without any re-training. The architecture was implemented on a robotic arm. Participants worked with the robotic arm to perform the task of picking toy car parts one by one and assembling the car while collaborating with the robot. Each participant repeated the task 3 times for each condition, Model or Wizard, in a random order. At the end of each trial participants completed a PeRDITA questionnaire. First, a test to rule out significant differences was performed, which yielded no significant results for any of the subjective and objective measures. As not having a significant difference does not necessarily mean similarity of conditions, to check for similarity, a Bayesian comparison of the conditions was performed next, which indicated a high probability of similarity between the model and Wizard performance. The high similarity to human-like performance observed for this more complex task supports the claim for the transferability of the models trained on a more generic task.

Game engines are increasingly used as simulation platforms by the autonomous vehicle community to develop vehicle control systems and test environments. A key requirement for simulation-based development and verification is determinism, since a deterministic process will always produce the same output given the same initial conditions and event history. Thus, in a deterministic simulation environment, tests are rendered repeatable and yield simulation results that are trustworthy and straightforward to debug. However, game engines are seldom deterministic. This paper reviews and identifies the potential causes and effects of non-deterministic behaviours in game engines. A case study using CARLA, an open-source autonomous driving simulation environment powered by Unreal Engine, is presented to highlight its inherent shortcomings in providing sufficient precision in experimental results. Different configurations and utilisations of the software and hardware are explored to determine an operational domain where the simulation precision is sufficiently high i.e. variance between repeated executions becomes negligible for development and testing work. Finally, a method of a general nature is proposed, that can be used to find the domains of permissible variance in game engine simulations for any given system configuration.

As autonomous robots move towards ubiquity, the need for robots to make decisions under risk that are trustworthy becomes increasingly significant; both to aid acceptance and to fully utilise their autonomous capabilities. We propose that incorporating a human approach to risk assessment into a robot's decision making process will increase user trust. This work investigates four robotic approaches to risk and, through a user study, explores the levels of trust placed in each. These approaches are: risk averse, risk seeking, risk neutral and a human approach to risk. Risk is artificially stimulated through performance-based compensation, in line with previous studies. The study was conducted in a virtual nuclear environment created using the Unity games engine. Forty participants were asked to complete a robot supervision task, in which they observed a robot making risk based decisions and were able to question the robot, question the robot further and ultimately accept or alter the robot's decision. It is shown that a robot that is risk seeking is significantly less trusted than a risk averse robot, a risk neutral robot and a robot utilising human approach to risk. There was found to be no significant difference between the levels of trust placed in the risk averse, risk neutral and human approach to risk. It is also found that the level to which participants question a robot's decisions does not form an accurate measure of trust. The results suggest that when designing a robot that must make risk based decisions during teleoperation in a hazardous environment, an engineer should avoid a risk seeking robot. However, that same engineer may choose whichever of the remaining risk profiles best suits the implementation, with knowledge that the trust in their system is unlikely to be significantly affected.

Future robotic assembly systems will allow product designers to upload their assembled product models remotely such that they can be assembled autonomously. In this work we present the architecture for a Click and Assemble Robotic Assembly (CARA) system. This architecture takes an assembly file uploaded by the user through a web interface as the only input. It then performs all the necessary planning before executing the assembly. To the authors' knowledge, this is the first time that all of the required components from previous advances in robotic assembly have been brought together, with all interconnecting challenges solved, into a complete working system for physical, real world assembly tasks. To demonstrate the implemented architecture capabilities, a real industrial product with tight tolerances was assembled from its CAD file only, illustrating the end to end robotic assembly premise in a real world setting.

In this paper, the use of a monitoring companion is proposed to more efficiently collect the process information during the tuning of industrial systems, and therefore to assist the system integrator in the optimization process for complex robotic installations. The monitoring companion consists of an industrial manipulator (monitoring robot) and a Unity-ROS framework for controlling it. We discuss in this paper the features that allow the solution to be re-used in different industrial application thanks to its compatibility with ROS. In particular, we present the approach based on admittance virtual fixtures and how we use such abstraction to track a moving target (it could be the end-effector of another robot for example). Moreover, the concept of varying compliance is introduced as a way to influence the motion of the monitoring robot on the virtual fixtures in the presence of obstacles. The experiments have been conducted in simulation as well as on real hardware to test the accuracy of the system at respecting the virtual fixtures with both a static and a moving monitoring target, although in the current implementation the varying compliance was not included in the experiments.

The recent progress made in the field of Augmented Reality/Mixed Reality (AR/MR) has opened new possibilities and approaches to research areas that can benefit from 3D visualization of digital content in the real world. In fact, human-robot interaction design and the design of user interfaces have very much to gain from MR technologies. Nonetheless, designing the user-robot interaction and processing multimodal feedbacks are very challenging tasks. In this paper we focus in particular on interactions in mixed reality.

Strategies are necessary to mitigate the impact of unexpected behavior in collaborative robotics, and research to develop solutions is lacking. Our aim here was to explore the benefits of an affective interaction, as opposed to a more efficient, less error prone but non-communicative one. The experiment took the form of an omelet-making task, with a wide range of participants interacting directly with BERT2, a humanoid robot assistant. Having significant implications for design, results suggest that efficiency is not the most important aspect of performance for users; a personable, expressive robot was found to be preferable over a more efficient one, despite a considerable trade off in time taken to perform the task. Our findings also suggest that a robot exhibiting human-like characteristics may make users reluctant to `hurt its feelings'; they may even lie in order to avoid this.

It has been frequently argued that the addition of haptic sensing to tele-operated surgical robots would benefit surgeon performance. Conventional haptic sensing technologies are impractical for application to minimally invasive surgery, due to size, sterilization robustness and cost vs. tool disposability. In this work we validate the concept of remote force measurement, where force interactions at the tip of a surgical tool are observed via simple torque sensors near the tool's actuators. This method provides reusable sensors located outside of the human body and so sidesteps many key issues that have limited practical haptic sensing in this scenario. Though such methods have been proposed and criticized by several groups in the past, we have been unable to locate quantitative results in the literature. Here, we provide initial remote force interaction measurements on a da Vinci EndoWrist needle driver tool. The measurements were obtained via simple custom torque sensors which easily attach to any EndoWrist tool. The complex cable-pulley transmission of the tool introduces expected nonlinearities over distal measurements. These effects are more pronounced in flexion than abduction DOFs. Despite these effects, the unprocessed torque data identifies contact with synthetic soft tissue at various actuator velocities and during external shaft loading.

We present a new approach to text line aggregation that can work as both a line formation stage for a myriad of text segmentation methods (over all orientations) and as an extra level of filtering to remove false text candidates. The proposed method is centred on the processing of candidate text components based on local and global measures. We use orientation histograms to build an understanding of paragraphs, and filter noise and construct lines based on the discovery of prominent orientations. Paragraphs are then reduced to seed components and lines are reconstructed around these components. We demonstrate results for text aggregation on the ICDAR 2003 Robust Reading Competition data, and also present results on our own more complex data set. Copyright © 2014 SCITEPRESS.

The human finger pad is known to be highly sensitive to lateral skin deformation, which is present during the palpation of soft objects containing hard lumps. This study investigates the effects of this lateral information on the accuracy and reliability of lump localization through a teletaction system. The results are expected to benefit future telesurgical robots for remote palpation. A previously proven deformation-based tactile feedback system was modified to include laterotactile feedback using a new version of a biologically-inspired fingertip sensor (TACTIP) and a remotely-actuated electromechanical shape display. These novel devices were mounted on to a newly developed teleoperated system equipped with visual and force feedback, through which test subjects palpated a number of soft artificial tissue samples. The level of lateral feedback was altered between trials, with constant visual, force, and normal-direction tactile feedback provided throughout. The surprising results show that the addition of lateral feedback offered no benefit to the subjects' ability to detect and localise embedded objects. Reasons for this observation are discussed. It is concluded that normal-direction tactile feedback is likely to be sufficient for lump detection even when lateral motions are made, although feedback from the subjects indicates that the addition benefited their perception of interactions with the system.

This paper represents the first steps in investigating issues emerging from a scenario where a robot interacts with a group of people around an interactive tabletop. In particular, the impact that a humanoid robot, acting as an avatar for a remote member of the group, has on the collaboration between the members of the group is investigated. This is carried out in the context where the avatar is depicting the actions being carried out by the remote user during a game. A preliminary study was performed to find out how the users interact when they are all co-located. The experiment is then carried out in two variants, in the first, a member of the group is moved to a remote location and allowed to interact with the other members using audio support alone. In the second, a robot is used to represent the actions of the remote user to the co-located users. The results obtained indicate that the addition of an avatar, used in such a way, has a positive impact on group interaction and allows for the task to be completed more successfully than when the avatar is not used. Further, comparative analysis of videos of the group behavior with a co-located team and group behavior with one member represented by the robot avatar, showed similar cooperative action.

A biomimetic mobile robot called “Shrewbot” has been built as part of a neuroethological study of the mammalian facial whisker sensory system. This platform has been used to further evaluate the problem space of whisker based tactile Simultaneous Localisation And Mapping (tSLAM). Shrewbot uses a biomorphic 3-dimensional array of active whiskers and a model of action selection based on tactile sensory attention to explore a circular walled arena sparsely populated with simple geometric shapes. Datasets taken during this exploration have been used to parameterise an approach to localisation and mapping based on probabilistic occupancy grids. We present the results of this work and conclude that simultaneous localisation and mapping is possible given only noisy odometry and tactile information from a 3-dimensional array of active biomimetic whiskers and no prior information of features in the environment.

The development of trustworthy human-assistive robots is a challenge that goes beyond the traditional boundaries of engineering. Essential components of trustworthiness are safety, predictability and usefulness. In this paper we demonstrate that the integration of joint action understanding from human-human interaction into the human-robot context can significantly improve the success rate of robot-to-human object handover tasks. We take a two layer approach. The first layer handles the physical aspects of the handover. The robot's decision to release the object is informed by a Hidden Markov Model that estimates the state of the handover. Inspired by human-human handover observations, we then introduce a higher-level cognitive layer that models behaviour characteristic for a human user in a handover situation. In particular, we focus on the inclusion of eye gaze / head orientation into the robot's decision making. Our results demonstrate that by integrating these non-verbal cues the success rate of robot-to-human handovers can be significantly improved, resulting in a more robust and therefore safer system.

Living organisms are complex systems, and yet they possess extremely high degrees of reliability. Since failures are local, their repair will often be taken on the local (cell) level. Engineers have long sought systems that could offer similar reliability and have relatively recently started trying to integrate ideas inspired by nature into the modern silicon technology of today. While bio-inspired proposals inspired by multicellular systems demonstrated feasibility, the resulting systems were often unduly complex. We are proposing a radically new methodology inspired by the characteristics, morphology, and behavior of simpler prokaryotic bacteria and bacterial communities. The hypothesis we use is that such simple unicellular organisms could help to build simpler cost effective systems, but with improved reliability than hitherto achieved by other methods. The result is a cellular array-based fault-tolerant electronic system with online self-test and self-repair capability. These ideas are simulated, tested, and verified through the successful construction of demonstrators: a proportional, integral, and differential and a robot controller. This paper discusses the underlying biological principles that guide our research and the bio-inspired model that we have derived. It also gives a detailed circuit and system description of the architecture and its run-time self-diagnostic and self-repair capability.

As elaborate human-robot interaction capabilities continue to develop, humans will increasingly be in proximity with robots, and the management of the ongoing control in case of breakdown becomes increasingly important: taking care of what happens when cooperation goes wrong. The current research addresses three categories of breakdowns where cooperation can go wrong. In the first category, the human detects some type of problem and generates a self-issued stop signal, with a physical palm up posture. In the second category, the human becomes distracted, and physically changes his orientation away from the shared space of cooperation. In the final category that we investigate, the human becomes physically close to the robot such that safety limits are reached and detected by the robot. In each of these three cases, the robot cognitive system detects the failure via the perception of distinct physical states from motion capture: the hand up posture; change in head orientation; and physical distance reaching a minimum threshold. In each case the robot immediately halts the current action. Then, the system should recover appropriately. Each error type returns a specific code, allowing the Supervisor system to handle the specific type of error. Our cognitive system allows the robot to learn composite actions, as a sequence of atomic actions. These composite actions can then be composed into higher level plans. When a plan fails at the level of a composite action, the recovery method is not trivial: should recovery take place at the level of the composite action, or the actual atomic action which physically failed? As the best recovery may depend on the physical context, we expand the plan into atomic actions, and recover at this level, allowing the user to specify whether the action should be skipped or retried. We demonstrate that this system allows graceful recovery from three principal categories of interaction breakdown, and provides an invaluable mechanism for preserving the integrity of cooperative HRI.

In this paper we have investigated the efficacy of applying multi-objective optimisation to Cartesian genetic programming (CGP) when used for evolution of cell-array configurations. A cell-array is a proposed type of custom FPGA, where digital circuits can be formed from interconnected configurable cells; thus, the CGP nodes are more complex than in its standard implementation. We have described modifications to a previously described optimisation algorithm that has led to significant improvements in performance; circuits close to a hand designed equivalent have been found, in terms of the optimised objectives. Additionally we have investigated the effect of circuit decomposition techniques on evolutionary performance. We found that using a hybrid of input and output decomposition techniques substantial reductions in evolution time were observed. Further, while the number of circuit inputs is the key factor for functional evolution time, the number of circuit outputs is the key factor for optimisation time.

This paper shows a successful application of evolutionary algorithms for the design and optimisation of complex real world digital circuit that is a 32-Step Traffic Lights Controller. It discusses two important features of electronic design through evolutionary processes; creativity and innovation. Results are compared to conventional design topologies; and attempt to analyse the evolved designs is presented.

Cooperative gestures are a key aspect of human-human pro-social interaction. Thus, it is reasonable to expect that endowing humanoid robots with the ability to use such gestures when interacting with humans would be useful. However, while people are used to responding to such gestures expressed by other humans, it is unclear how they might react to a robot making them. To explore this topic, we conducted a within-subjects, video based laboratory experiment, measuring time to cooperate with a humanoid robot making interactional gestures. We manipulated the gesture type (beckon, give, shake hands), the gesture style (smooth, abrupt), and the gesture orientation (front, side). We also employed two measures of individual differences: negative attitudes toward robots (NARS) and human gesture decoding ability (DANVA2-POS). Our results show that people cooperate with abrupt gestures more quickly than smooth ones and front-oriented gestures more quickly than those made to the side, people's speed at decoding robot gestures is correlated with their ability to decode human gestures, and negative attitudes toward robots is strongly correlated with a decreased ability in decoding human gestures.

An adaptive multi-dimensional compliance model reference controller was implemented in real-time on a 4 degrees of freedom (DOF) of the humanoid Bristol-Elumotion-Robotic-Torso II (BERT II) arm in Cartesian space. The robot manipulator has been controlled in such a way as to follow the compliant passive behaviour of a reference mass-spring-damper system model subject to externally sensed forces/torques in all DOF. The relevant reference model converts all measured torques into their equivalent forces at the end-effector and reacts accordingly. The suggested control scheme takes in particular account of the multi-variable aspect and the problem of body own torques when measuring external torques. The redundant DOF were used to control the robot motion in a human-like pattern via effort minimization. Associated actuator saturation issues were addressed by incorporating a novel anti-windup (AW) compensator.

Sensory signals are often caused by one's own active movements. This raises a problem of discriminating between self-generated sensory signals and signals generated by the external world. Such discrimination is of general importance for robotic systems, where operational robustness is dependent on the correct interpretation of sensory signals. Here, we investigate this problem in the context of a whiskered robot. The whisker sensory signal comprises two components: one due to contact with an object (externally generated) and another due to active movement of the whisker (self-generated). We propose a solution to this discrimination problem based on adaptive noise cancelation, where the robot learns to predict the sensory consequences of its own movements using an adaptive filter. The filter inputs (copy of motor commands) are transformed by Laguerre functions instead of the often-used tapped-delay line, which reduces model order and, therefore, computational complexity. Results from a contact-detection task demonstrate that false positives are significantly reduced using the proposed scheme.

This paper discusses two important features of electronic design through evolutionary processes; creativity and innovation. Hence, conventional design methodologies are discussed and compared with their counterparts via evolutionary processes. An evolutionary search is used as an engine for discovering new designs for a real world application. Attempts to extract some useful principles from the evolved designs are presented and results are compared to conventional design topologies for the same problems.

In order for anthropomorphic robots to communicate with people in a human-like way they need to produce gestures along with speech. Beat gestures are a key category of gestures, accompanying emphasis and timing of talk. The standard approach for autonomously adding gestures to speech for artificial agents to perform, is to identify when gestures should occur, and then select appropriate movements from a prewritten library. However, selecting naively from a library is unlikely to result in particularly human-like gesture sequences. In order to obtain examples of engaging human beat gestures, we studied videos of chat show hosts. Our analysis reveals rules for the creation of human-like beat gestures. We have combined these rules with hand scripted non-beat gestures, to produce a monologue with a complete set of accompanying gestures; it is designed for performance by our humanoid robot BERTI. In order to test the gestures produced we carried out a pilot study at the London Science Museum. The results of the study indicate that having correctly designed gesture sequences improves observer engagement.

The human sciences have demonstrated that gesture is a critical element of human communication. While existing graphical solutions are appropriate for virtual agents, solving arm trajectories for physically embodied robots requires that we consider the challenges of robot dynamics within a realtime gesture framework. We explore and evaluate a low computational-cost gesture production algorithm that can generate adequate gesture trajectories in a humanoid torso, as judged by participants in Human-Robot gesturing studies presented in this paper. Our approach produces a constrained inverse-kinematic solution for the start and end points, and generates appropriate wrist angles. Gesture time is used to calculate the joint accelerations to give a smooth, direct hand movement. Selecting open hand gestures as an example gesture sub-domain, we implement our controller on BERTI, a bespoke upper-torso humanoid robot (Fig. 2). A qualitative pilot study highlights gesture features salient to users: gesture shape, timing, naturalness and smoothness. A controlled experimental study then demonstrates that, by these metrics, our algorithm performs well; despite some dissimilarities with users' own gestures. We establish some salient points of robot gestures based on these studies.

In this paper, a model of cerebellar function is implemented and evaluated in the control of a robot eye actuated by pneumatic artificial muscles. The investigated control problem is stabilization of the visual image in response to disturbances. This is analogous to the vestibuloocular reflex (VOR) in humans. The cerebellar model is structurally based on the adaptive filter, and the learning rule is computationally analogous to least-mean squares, where parameter adaptation at the parallel fiber/Purkinje cell synapse is driven by the correlation of the sensory error signal (carried by the climbing fiber) and the motor command signal. Convergence of the algorithm is first analyzed in simulation on a model of the robot and then tested online in both one and two degrees of freedom. The results show that this model of neural function successfully works on a real-world problem, providing empirical evidence for validating: 1) the generic cerebellar learning algorithm; 2) the function of the cerebellum in the VOR; and 3) the signal transmission between functional neural components of the VOR.

Embryonic hardware systems satisfy the fundamental characteristics found in nature which contribute to the development of any multi-cellular living being. Attempts of researchers' in this field to learn from nature have yielded promising results; they proved the feasibility of applying nature-like mechanisms to the world of digital electronics with self-diagnostic and self-healing characteristics, Design by humans however often results in very complex hardware architectures, requiring a large amount of manpower and computational resources. A wider objective is to find novel solutions to design such complex architectures for Embryonic Systems, by problem decomposition and unique design methodologies so that system functionality and performance will not be compromised. Design automation using reconfigurable hardware and EA (evolutionary algorithm), such as GA (genetic algorithms), is one way to tackle this issue. This concept applies the notion of EHW (evolvable hardware) to the problem domain. Unlocking the power of EHW for both novel design solutions and for circuit optimisation has attracted many researchers since the early '90s. The promise of using genetic algorithms through evolvable hardware design will, in this paper, be demonstrated by the authors by evolving a relatively simple combinatorial logic circuit (full-adder).

We present an experimental comparison between two approaches to optimization of the rules for a fuzzy controller. More specifically, the problem is autonomous acquisition of an “investigative” obstacle avoidance competency for a mobile robot. We report on results from investigating two alternative approaches to the use of a Learning Classifier System (LCS) to optimize the fuzzy rule base. One approach operates at the level of whole rule bases, the “Pittsburgh” LCS. The other approach operates at the level of individual rules, the “Michigan” LCS. In this work, both of these Fuzzy Classifier Systems were designed to operate only on the rules of fuzzy controllers, with predefined fuzzy membership functions. There are two main results from this work. First, both approaches were capable of producing fuzzy controllers with subtle interactions between rules leading to competencies exceeding that of the hand-coded fuzzy controller presented in this article. Second, the Michigan approach suffered more seriously than the Pittsburgh approach from the well-known LCS “cooperation/competition” problem, which is accentuated here by the structural combination of Evolutionary Computation and a fuzzy system. This problem was alleviated a little by the combination of a clustered subpopulation niche system and a fitness-sharing scheme applied to the Michigan approach, but still remains. © 2007 Wiley Periodicals, Inc. Int J Int Syst 22: 993–1019, 2007.

In this paper, we present two versions of a hardware processing architecture for modeling large networks of leaky-integrate-and-fire (LIF) neurons; the second version provides performance enhancing features relative to the first. Both versions of the architecture use fixed-point arithmetic and have been implemented using a single field-programmable gate array (FPGA). They have successfully simulated networks of over 1000 neurons configured using biologically plausible models of mammalian neural systems. The neuroprocessor has been designed to be employed primarily for use on mobile robotic vehicles, allowing bio-inspired neural processing models to be integrated directly into real-world control environments. When a neuroprocessor has been designed to act as part of the closed-loop system of a feedback controller, it is imperative to maintain strict real-time performance at all times, in order to maintain integrity of the control system. This resulted in the reevaluation of some of the architectural features of existing hardware for biologically plausible neural networks (NNs). In addition, we describe a development system for rapidly porting an underlying model (based on floating-point arithmetic) to the fixed-point representation of the FPGA-based neuroprocessor, thereby allowing validation of the hardware architecture. The developmental system environment facilitates the cooperation of computational neuroscientists and engineers working on embodied (robotic) systems with neural controllers, as demonstrated by our own experience on the Whiskerbot project, in which we developed models of the rodent whisker sensory system.

Both a FPGA implementation of a rodent's tactile sensory system and a neural FPGA based hardware processor, mimicking the brainstem behaviour are presented. They have principally been designed using biological considerations. The two systems are being ported on a single FPGA platform and will ultimately be embedded on a mobile robot, which will operate in real world environments for object recognition and surface textural discrimination purposes

The implementation of a tactile sensory system for neuromorphic signal processing applications is presented. It has been developed by modelling the structure and behaviour of real rodent facial vibrissae. The primary afferents have been modelled using empirical data taken from electrophysiological measurements, and implemented in real time using hardware processors (DSP/FPGA). Pipelining techniques were employed to maximise the utility of the FPGA hardware. The system is to be integrated into a more complete whisker sensory model, including neural structures within the central nervous system, which can be used to orientate a mobile robot

Electronic systems, no matter how clever and intelligent they are, cannot yet demonstrate the reliability that biological systems can. Perhaps we can learn from these processes, which have developed through millions of years of evolution, in our pursuit of highly reliable systems. This article discusses how such systems, inspired by biological principles, might be built using simple embryonic cells. We illustrate how they can monitor their own functional integrity in order to protect themselves from internal failure or from hostile environmental effects and how faults caused by DNA mutation or cell death can be repaired and thus full system functionality restored.

The implementation of a large scale, leaky-integrate-and-fire neural network processor using the Xilinx Virtex-II family of field programmable gate array (FPGA) is presented. The processor has been designed to model biologically plausible networks of spiking neurons in real-time to assist with the control of a mobile robot. The real-time constraint has led to a re-evaluation of some of the established architectural and algorithmic features of previous spiking neural network based hardware. The design was coded and simulated using Handel-C hardware description language (HDL) and the DK3 design suite from Celoxica. The processor has been physically implemented and tested on a RC200 development board, also from Celoxica.

In this article, we describe some of the important currently used methods for solving classification problems, focusing on feature selection and extraction as parts of the overall classification task. We then go on to discuss likely future directions for research in this area, in the context of the other articles from this special issue. We propose that the next major step is the elaboration of a theory of how the methods of selection and extraction interact during the classification process for particular problem domains, along with any learning that may be part of the algorithms. Preferably this theory should be tested on a set of well-established benchmark challenge problems. Using this theory, we will be better able to identify the specific combinations that will achieve best classification performance for new tasks.

This paper describes a fresh interpretation of, and experimental modifications to, an architecture which has arisen from the author's previous work (1997). The previous published work concentrates upon the learning structure adopted, which is based on an adaptive heuristic critic. This paper focuses upon the nature of the knowledge representation acquired by the architecture, and in particular on the case of "latent" or "reward-free" learning. The purpose of this investigation is to show that our architecture can perform latent learning, and that this knowledge can be used to improve the performance of subsequent reward-based learning phases. The results of simulation experiments which have been devised to test performance under these circumstances are given.

Global optimization approaches to path planning for kinematically redundant manipulators, involving the optimization of integral cost criteria have been shown to exhibit certain advantages compared with local optimizers based on instantaneous criteria. We present some initial results of an approximation approach based on the direct search evolution strategy paradigm with specific modifications to encourage discovery of global optima and to explore in spaces with non-simply-connected topologies. In our problem formulation the path of the end effector is unconstrained, except at the end points, and consequently there exist an infinite number of possible hand paths, from which we seek that which gives the global minimum of an integral squared jerk cost functional.