Artificial neural networks (ANNs) are powerful methods for the classification of multi-dimensional data as well as for the control of dynamic systems. In general terms, ANNs consist of neurons that are, e.g., arranged in layers and interconnected by real-valued or binary neural couplings or weights. ANNs try mimicking the processing taking place in biological brains. The classification and generalization capabilities of ANNs are given by the interconnection architecture and the coupling strengths. To perform a certain classification or control task with a particular ANN architecture (i.e., number of neurons, number of layers, etc.), the inter-neuron couplings and their accordant coupling strengths must be determined (1) either by a priori design (i.e., manually) or (2) using training algorithms such as error back-propagation. The more complex the classification or control task, the less obvious it is how to determine an a priori design of an ANN, and, as a consequence, the architecture choice becomes somewhat arbitrary. Furthermore, rather than being able to determine for a given architecture directly the corresponding coupling strengths necessary to perform the classification or control task, these have to be obtained/learned through training of the ANN on test data. We report on the use of a Stochastic Optimization Framework (SOF; Fink, SPIE 2008) for the autonomous self-configuration of Artificial Neural Networks (i.e., the determination of number of hidden layers, number of neurons per hidden layer, interconnections between neurons, and respective coupling strengths) for performing classification or control tasks. This may provide an approach towards cognizant and self-adapting computing architectures and systems.

The Mission Operations Group at UC Berkeley's Space Sciences Laboratory operates a highly automated ground station and presently a fleet of seven satellites, each with its own associated command and control console. However, the requirement for prompt anomaly detection and resolution is shared commonly between the ground segment and all spacecraft. The efficient, low-cost operation and "lights-out" staffing of the Mission Operations Group requires that controllers and engineers be notified of spacecraft and ground system problems around the clock. The Berkeley Emergency Anomaly and Response System (BEARS) is an in-house developed web- and paging-based software system that meets this need. BEARS was developed as a replacement for an existing emergency reporting software system that was too closedsource, platform-specific, expensive, and antiquated to expand or maintain. To avoid these limitations, the new system design leverages cross-platform, open-source software products such as MySQL, PHP, and Qt. Anomaly notifications and responses make use of the two-way paging capabilities of modern smart phones.

Space solar applications will require PV modules with large Wp/kg values with stable output characteristics under extreme insolation conditions. This report is focused on the performance of Flexible PV (FPV) modules with moderate Wp/kg ratings when exposed to sustained insolation conditions reaching a maximum of 910 Watt/m². Three different FPV technologies have been considered, namely mono-crystalline silicon (c-Si), poly-crystalline silicon (pc-Si), and amorphous silicon (a-Si). Based on outdoor observations, the technology impact is most evident from the fill factor (FF) range demonstrated by the PV modules which is high (60 - 70%) for crystalline and polycrystalline silicon PV modules and moderate (50 - 60%) for amorphous silicon PV modules. A one diode model has been applied to all three PV modules and the theoretically calculated diode non ideality factor (n) has been compared with experimentally observed non idealities in terms of FF. Again the impact of technology is evident from 'n' value calculations which are 1.42 for c- Si, 2 for pc-Si and 3.7 for a-Si FPV modules.

At Caltech's Visual and Autonomous Exploration Systems Research Laboratory (http://autonomy.caltech.edu) an outdoor multi-rover testbed has been developed that allows for near real-time interactive or automatic control from anywhere in the world via the Internet. It enables the implementation, field-testing, and validation of algorithms/software and strategies for navigation, exploration, feature extraction, anomaly detection, and target prioritization with applications in planetary exploration, security surveillance, reconnaissance of disaster areas, military reconnaissance, and delivery of lethal force such as explosives for urban warfare. Several rover platforms have been developed, enabling testing of cooperative multi-rover scenarios (e.g., inter-rover communication/coordination) and distributed exploration of operational areas.

In this paper, the mechanical structure, dynamic model and control strategy of an omni-directional rolling spherical robot with a telescopic manipulator (BYQ-IV) are discussed in particular. The structure of the whole robot is included of the motion driving part, the manipulator part and the stability maintain part. The simplified dynamic model of the motion driving part is formed by the Kane method. Moreover, the distribute control system of the robot based on ARM processor and wireless communication system are introduced and the software architecture of control system is analyzed. This robot is designed for territory or lunar exploration. It not only has features like straight line motion, circular motion, zero turning radius and obstacle avoidance, but also is able to accomplish tasks such as stably grabbing and delivering assemblies. The experiment shows that the prototype of the spherical robot with telescopic manipulator can stably grasp a static target and carry it to a new location.

For realizing omni-directional movement and operating task of spherical space robot system, this paper describes an innovated prototype and analyzes dynamic characteristics of a spherical rolling robot with telescopic manipulator. Based on the Newton-Euler equations, the kinematics and dynamic equations of the spherical robot's motion are instructed detailedly. Then the motion simulations of the robot in different environments are developed with ADAMS. The simulation results validate the mathematics model of the system. And the dynamic model establishes theoretical basis for the latter job.

The underwater spherical robot has a spherical pressure hull which contains power modules, sensors, and so on. It lacks robot arms or end effectors but is highly maneuverable, for the simplest symmetrical geometry is the sphere. This paper analyzes the spherical robot's hydrodynamic model with CFD software, concludes the spherical robot's hydrodynamic characteristics, and compares these characteristics with the hydrodynamic model of another underwater robot which has a streamlined hull. The effect of sphere hydraulic resistance on the control of the robot is analyzed with some examples.

Part of the requirements of the future Constellation program is to optimize lunar surface operations and reduce hazards to astronauts. Toward this end, many robotic platforms, rovers in specific, are being sought to carry out a multitude of missions involving potential EVA sites survey, surface reconnaissance, path planning and obstacle detection and classification. 3D imaging lidar technology provides an enabling capability that allows fast, accurate and detailed collection of three-dimensional information about the rover's environment. The lidar images the region of interest by scanning a laser beam and measuring the pulse time-of-flight and the bearing. The accumulated set of laser ranges and bearings constitutes the threedimensional image. As part of the ongoing NASA Ames research center activities in lunar robotics, the utility of 3D imaging lidar was evaluated by testing Optech's ILRIS-3D lidar on board the K-10 Red rover during the recent Human - Robotics Systems (HRS) field trails in Lake Moses, WA. This paper examines the results of the ILRIS-3D trials, presents the data obtained and discusses its application in lunar surface robotic surveying and scouting.

The IR-ECD^TM (Infra-Red ElectroChromic Device) variable emitance device (VED) is an all-solid-state monolithic vacuum deposited thin film system with a unique metamaterial IR transparent-electrode system which functions as an electrically controlled dimmable mirror in the IR region. The maximum reflectance corresponding to the bleached condition of the system is around 90% (low-e condition, e=0.1). The minimum reflectance reaches nearly zero in the colored condition of the system (high emittance, e=1). The average emissivity modulation of the IRECD^TM is 0.7 in the 8-12 micron region, and at 9.7 micron (room temperature) it reaches a value of 0.9. Half and full emissivity modulations occur within 2 and10 minutes respectively. Because of its light weight (5g/m²), low voltage requirement (+/- 1 Volts), extremely good emissivity control properties (from 0 to 0.9 at 300K) and highly repeatable deposition process, the IR-ECD^TM technology is very attractive for satellite thermal control applications. The IR-ECD^TM has been under evaluation in a real space environment since March 8, 2007. This paper presents recent achievements of the IR-ECD^TM including space test results.

In Space & Defense fields, there is a trend for miniaturisation in active optics, fine instruments, robotic missions, microsatellites, UAVs, MAVs which directly impact on the design of actuators. A new generation of small and smart actuators such like piezoelectric (piezo) actuators, are responding to this trend, thanks to their capacity to offer high energy density and to support both extreme and various requirements. In Space vehicles, UAVs, missiles, military vehicles, etc., onboard place and available electric power can be very limited. For instance, a micro satellite often must operate all its instruments with less than 100W of power. As a result, allocated electric power per actuator is typically between 0.1 to 10W. This is also the case in small UAVs and in MAVs. Because of the high cost of embedded mass, space & military actuators need also to offer high output energy to mass ratio. One of the main difficulties is often the ability to withstand launching vibrations and shocks. Space environments add other constrains. A clear example is the vacuum conditions, which can induce difficulties to release the heat out off the actuator or for out gassing near optics. Other critical spacerelated environmental conditions include the thermal operation range required as well as the radiation-resistant requirements. In other situations, actuator strength to humidity is often an issue, especially for piezoelectric ceramics. Thus, the success of the application relies not only on design issues but also on material reliability. Specific actions at this level are needed to be undertaken to secure space projects. To cope with these issues and to illustrate the trend, the piezo actuators and mechanisms from Cedrat are presented. They have been initially developed and qualified to meet space requirements but logically found also applications in defense and micro aerial vehicle fields, for various micromechatronic functions. The paper presents typical applications and piezo mechatronic based system such like, piezo micro-scanning stage for IR camera resolution enhancement, piezo active flap on helicopter blade for noise reduction, micro amplified piezo actuator for tilting MAV rotor, hollow piezo actuator for external laser cavity tuning of a space LIDAR, in order to discuss the state-of-the-art performance and deduce further needs.