Mutual Information and Perplexity based Clustering of Dialogue Information for Dynamic Adaptation of Language Models

We present two approaches to cluster dialogue-based information obtained by the speech understanding module and the dialogue manager of a spoken dialogue system. The purpose is to estimate a language model related to each cluster, and use them to dynamically modify the model of the speech recognizer at each dialogue turn. In the first approach we build the cluster tree using local decisions based on a Maximum Normalized Mutual Information criterion. In the second one we take global decisions, based on the optimization of the global perplexity of the combination of the cluster-related LMs. Our experiments show a relative reduction of the word error rate of 15.17%, which helps to improve the performance of the understanding and the dialogue manager modules.

Mining recurring concepts in a dynamic feature space

Most data stream classification techniques assume that the underlying feature space is static. However, in real-world applications the set of features and their relevance to the target concept may change over time. In addition, when the underlying concepts reappear, reusing previously learnt models can enhance the learning process in terms of accuracy and processing time at the expense of manageable memory consumption. In this paper, we propose mining recurring concepts in a dynamic feature space (MReC-DFS), a data stream classification system to address the challenges of learning recurring concepts in a dynamic feature space while simultaneously reducing the memory cost associated with storing past models. MReC-DFS is able to detect and adapt to concept changes using the performance of the learning process and contextual information. To handle recurring concepts, stored models are combined in a dynamically weighted ensemble. Incremental feature selection is performed to reduce the combined feature space. This contribution allows MReC-DFS to store only the features most relevant to the learnt concepts, which in turn increases the memory efficiency of the technique. In addition, an incremental feature selection method is proposed that dynamically determines the threshold between relevant and irrelevant features. Experimental results demonstrating the high accuracy of MReC-DFS compared with state-of-the-art techniques on a variety of real datasets are presented. The results also show the superior memory efficiency of MReC-DFS.

A Monitoring Infrastructure for FPGA Self-Awareness and Dynamic Adaptation

Variabilities associated with CMOS evolution affect the yield and performance of current digital designs. FPGAs, which are widely used for fast prototyping and implementation of digital circuits, also suffer from these issues. Proactive approaches start to appear to achieve self-awareness and dynamic adaptation of these devices. To support these techniques we propose the employment of a multi-purpose sensor network. This infrastructure, through adequate use of configuration and automation tools, is able to obtain relevant data along the life cycle of an FPGA. This is realised at a very reduced cost, not only in terms of area or other limited resources, but also regarding the design effort required to define and deploy the measuring infrastructure. Our proposal has been validated by measuring inter-die and intra-die variability in different FPGA families.

AUTOFLY-Aid: Flight Deck Automation Support with Dynamic 4D Trajectory Management for Responsive and Adaptive Airborne Collision Avoidance

AUTOFLY-Aid Project aims to develop and demonstrate novel automation support algorithms and tools to the flight crew for flight critical collision avoidance using “dynamic 4D trajectory management”. The automation support system is envisioned to improve the primary shortcomings of TCAS, and to aid the pilot through add-on avionics/head-up displays and reality augmentation devices in dynamically evolving collision avoidance scenarios. The main theoretical innovative and novel concepts to be developed by AUTOFLY-Aid project are a) design and development of the mathematical models of the full composite airspace picture from the flight deck’s perspective, as seen/measured/informed by the aircraft flying in SESAR 2020, b) design and development of a dynamic trajectory planning algorithm that can generate at real-time (on the order of seconds) flyable (i.e. dynamically and performance-wise feasible) alternative trajectories across the evolving stochastic composite airspace picture (which includes new conflicts, blunder risks, terrain and weather limitations) and c) development and testing of the Collision Avoidance Automation Support System on a Boeing 737 NG FNPT II Flight Simulator with synthetic vision and reality augmentation while providing the flight crew with quantified and visual understanding of collision risks in terms of time and directions and countermeasures.

Power management techniques in an FPGA-Based WSN node for high performance application

In this work, the power management techniques implemented in a high-performance node for Wireless Sensor Networks (WSN) based on a RAM-based FPGA are presented. This new node custom architecture is intended for high-end WSN applications that include complex sensor management like video cameras, high compute demanding tasks such as image encoding or robust encryption, and/or higher data bandwidth needs. In the case of these complex processing tasks, yet maintaining low power design requirements, it can be shown that the combination of different techniques such as extensive HW algorithm mapping, smart management of power islands to selectively switch on and off components, smart and low-energy partial reconfiguration, an adequate set of save energy modes and wake up options, all combined, may yield energy results that may compete and improve energy usage of typical low power microcontrollers used in many WSN node architectures. Actually, results show that higher complexity tasks are in favor of HW based platforms, while the flexibility achieved by dynamic and partial reconfiguration techniques could be comparable to SW based solutions.

Predicting academic performance with learning analytics in virtual learning environments: A comparative study of three interaction classifications

Learning analytics is the analysis of static and dynamic data extracted from virtual learning environments, in order to understand and optimize the learning process. Generally, this dynamic data is generated by the interactions which take place in the virtual learning environment. At the present time, many implementations for grouping of data have been proposed, but there is no consensus yet on which interactions and groups must be measured and analyzed. There is also no agreement on what is the influence of these interactions, if any, on learning outcomes, academic performance or student success. This study presents three different extant interaction typologies in e-learning and analyzes the relation of their components with students? academic performance. The three different classifications are based on the agents involved in the learning process, the frequency of use and the participation mode, respectively. The main findings from the research are: a) that agent-based classifications offer a better explanation of student academic performance; b) that at least one component in each typology predicts academic performance; and c) that student-teacher and student-student, evaluating students, and active interactions, respectively, have a significant impact on academic performance, while the other interaction types are not significantly related to academic performance.

Measuring dynamic effects on underpasses of high-speed railway lines

Underpasses are common in modern railway lines. Wildlife corridors and drainage conduits often fall into this category of partially buried structures. Their dynamic behaviour has received far less attention than that of other structures such as bridges, but their large number makes their study an interesting challenge in order to achieve safe and cost-effective structures. As ballast operations are a key life cycle cost, and excessive vibrations increase the need of ballast regulation in order to ensure track geometry, special attention is paid to accelerations, the values of which should be limited to avoid track instability according to Eurocode. In this paper, the data obtained during on site measurements on culverts belonging to a Spanish high-speed train line are presented. A set of six rectangular-shaped, closed-frame underpasses were monitored under traffic loading. Acceleration records at different points of the structures are presented and discussed. They reveal a non-uniform dynamic response of the roof-slab, with the highest observed values below the occupied track. Also, they indicate that the dynamic response is important up to frequencies higher than those usually observed for standard simply supported bridges. Finally, they are used to obtain a heuristic rule to estimate acceleration levels on the roof-slab.

Dynamic Soil-Structure Interaction in Bridge Abutments

Strong motion obtained in instrumental short-span bridges show the importance of the abutments in the dynamic response of the whole structure. Many models have been used in order to take into account the influence of pier foundations although no reliable ones have been used to analyse the abutment performance. In this work three-dimensional Boundary Element models in frequency domain have been proposed and dimensionless dynamic stiffness of standard bridge abutments have been obtained.

On the dynamic adaptation of language models based on dialogue information

We present an approach to adapt dynamically the language models (LMs) used by a speech recognizer that is part of a spoken dialogue system. We have developed a grammar generation strategy that automatically adapts the LMs using the semantic information that the user provides (represented as dialogue concepts), together with the information regarding the intentions of the speaker (inferred by the dialogue manager, and represented as dialogue goals). We carry out the adaptation as a linear interpolation between a background LM, and one or more of the LMs associated to the dialogue elements (concepts or goals) addressed by the user. The interpolation weights between those models are automatically estimated on each dialogue turn, using measures such as the posterior probabilities of concepts and goals, estimated as part of the inference procedure to determine the actions to be carried out. We propose two approaches to handle the LMs related to concepts and goals. Whereas in the first one we estimate a LM for each one of them, in the second one we apply several clustering strategies to group together those elements that share some common properties, and estimate a LM for each cluster. Our evaluation shows how the system can estimate a dynamic model adapted to each dialogue turn, which helps to improve the performance of the speech recognition (up to a 14.82% of relative improvement), which leads to an improvement in both the language understanding and the dialogue management tasks.

Localization and Tracking in Energy-Constrained Networks with Dynamic Connectivity

Systems used for target localization, such as goods, individuals, or animals, commonly rely on operational means to meet the final application demands. However, what would happen if some means were powered up randomly by harvesting systems? And what if those devices not randomly powered had their duty cycles restricted? Under what conditions would such an operation be tolerable in localization services? What if the references provided by nodes in a tracking problem were distorted? Moreover, there is an underlying topic common to the previous questions regarding the transfer of conceptual models to reality in field tests: what challenges are faced upon deploying a localization network that integrates energy harvesting modules? The application scenario of the system studied is a traditional herding environment of semi domesticated reindeer (Rangifer tarandus tarandus) in northern Scandinavia. In these conditions, information on approximate locations of reindeer is as important as environmental preservation. Herders also need cost-effective devices capable of operating unattended in, sometimes, extreme weather conditions. The analyses developed are worthy not only for the specific application environment presented, but also because they may serve as an approach to performance of navigation systems in absence of reasonably accurate references like the ones of the Global Positioning System (GPS). A number of energy-harvesting solutions, like thermal and radio-frequency harvesting, do not commonly provide power beyond one milliwatt. When they do, battery buffers may be needed (as it happens with solar energy) which may raise costs and make systems more dependent on environmental temperatures. In general, given our problem, a harvesting system is needed that be capable of providing energy bursts of, at least, some milliwatts. Many works on localization problems assume that devices have certain capabilities to determine unknown locations based on range-based techniques or fingerprinting which cannot be assumed in the approach considered herein. The system presented is akin to range-free techniques, but goes to the extent of considering very low node densities: most range-free techniques are, therefore, not applicable. Animal localization, in particular, uses to be supported by accurate devices such as GPS collars which deplete batteries in, maximum, a few days. Such short-life solutions are not particularly desirable in the framework considered. In tracking, the challenge may times addressed aims at attaining high precision levels from complex reliable hardware and thorough processing techniques. One of the challenges in this Thesis is the use of equipment with just part of its facilities in permanent operation, which may yield high input noise levels in the form of distorted reference points. The solution presented integrates a kinetic harvesting module in some nodes which are expected to be a majority in the network. These modules are capable of providing power bursts of some milliwatts which suffice to meet node energy demands. The usage of harvesting modules in the aforementioned conditions makes the system less dependent on environmental temperatures as no batteries are used in nodes with harvesters--it may be also an advantage in economic terms. There is a second kind of nodes. They are battery powered (without kinetic energy harvesters), and are, therefore, dependent on temperature and battery replacements. In addition, their operation is constrained by duty cycles in order to extend node lifetime and, consequently, their autonomy. There is, in turn, a third type of nodes (hotspots) which can be static or mobile. They are also battery-powered, and are used to retrieve information from the network so that it is presented to users. The system operational chain starts at the kinetic-powered nodes broadcasting their own identifier. If an identifier is received at a battery-powered node, the latter stores it for its records. Later, as the recording node meets a hotspot, its full record of detections is transferred to the hotspot. Every detection registry comprises, at least, a node identifier and the position read from its GPS module by the battery-operated node previously to detection. The characteristics of the system presented make the aforementioned operation own certain particularities which are also studied. First, identifier transmissions are random as they depend on movements at kinetic modules--reindeer movements in our application. Not every movement suffices since it must overcome a certain energy threshold. Second, identifier transmissions may not be heard unless there is a battery-powered node in the surroundings. Third, battery-powered nodes do not poll continuously their GPS module, hence localization errors rise even more. Let's recall at this point that such behavior is tight to the aforementioned power saving policies to extend node lifetime. Last, some time is elapsed between the instant an identifier random transmission is detected and the moment the user is aware of such a detection: it takes some time to find a hotspot. Tracking is posed as a problem of a single kinetically-powered target and a population of battery-operated nodes with higher densities than before in localization. Since the latter provide their approximate positions as reference locations, the study is again focused on assessing the impact of such distorted references on performance. Unlike in localization, distance-estimation capabilities based on signal parameters are assumed in this problem. Three variants of the Kalman filter family are applied in this context: the regular Kalman filter, the alpha-beta filter, and the unscented Kalman filter. The study enclosed hereafter comprises both field tests and simulations. Field tests were used mainly to assess the challenges related to power supply and operation in extreme conditions as well as to model nodes and some aspects of their operation in the application scenario. These models are the basics of the simulations developed later. The overall system performance is analyzed according to three metrics: number of detections per kinetic node, accuracy, and latency. The links between these metrics and the operational conditions are also discussed and characterized statistically. Subsequently, such statistical characterization is used to forecast performance figures given specific operational parameters. In tracking, also studied via simulations, nonlinear relationships are found between accuracy and duty cycles and cluster sizes of battery-operated nodes. The solution presented may be more complex in terms of network structure than existing solutions based on GPS collars. However, its main gain lies on taking advantage of users' error tolerance to reduce costs and become more environmentally friendly by diminishing the potential amount of batteries that can be lost. Whether it is applicable or not depends ultimately on the conditions and requirements imposed by users' needs and operational environments, which is, as it has been explained, one of the topics of this Thesis.

T-REX: Bare electro-dynamic tape-tether technology experimetn on sounding rocket S520

The project to verify the performance of space tether technology was successfully demonstrated by the launch of the sounding rocket S520 the 25tu. The project is the space demonstration of science and engineering technologies of a bare tape electrodynamic tether (EDT) in the international campaign between Japan, USA, Europe and Australia. Method of "Inverse ORIGAMI (Tape tether folding)" was employed in order to deploy the bare tape EDT in a short period time of the suborbital flight. The deployment of tape tether was tested in a various experimental schemes on ground to show high reliability of tape tether deployment. The rocket was launched on the summer of 2010 and deployed a bare electro-dynamic tape tether with length 132.6 m, which is the world record of the length deployment of tape tether. The verification of tether technology has found a variety kind of science and technology results as the first in the humankind and will lead a large number of applications of space tether technologies

Analysis of Proseds bare-tether performance

NASA's tether experiment ProSEDS will be placed in orbit on board a Delta-II rocket in early 2003. ProSEDS will test bare-tether electron collection, deorbiting of the rocket second stage, and the system dynamic stability. ProSEDS performance will vary both because ambient conditions change along the orbit and because tether-circuit parameters follow a step by step sequence in the current operating cycle. In this work we discuss how measurements of tether current and bias, plasma density, and deorbiting rate can be used to check the OML law for current collection. We review circuit bulk elements; characteristic lengths and energies that determine collection (tether radius, electron thermal gyroradius and Debye length, particle temperatures, tether bias, ion ram energy); and lengths determining current and bias profiles along the tether (extent of magnetic self-field, a length gauging ohmic versus collection impedances, tether length). The analysis serves the purpose of estimating ProSEDS behavior in orbit and fostering our ability for extrapolating ProSEDS flight data to different tether and environmental conditions.

Performance-based seismic design of a retaining wall

The design of containment walls suffering seismic loads traditionally has been realized with methods based on pseudoanalitic procedures such as Mononobe-Okabe's method, which it has led in certain occasions to insecure designs, that they have produced the ruin of many containment walls suffering the action of an earthquake. The recommendations gathered in Mononobe-Okabe's theory have been included in numerous Codes of Seismic Design. It is clear that a revision of these recommendations must be done. At present there is taking place an important review of the design methods of anti-seismic structures such as containment walls placed in an area of numerous earthquakes, by means of the introduction at the beginning of the decade of 1990 the Displacement Response Spectrum (DRS) and the Capacity Demand Diagram (CDD) that suppose an important change in the way of presenting the Elastic Response Spectrum (ERS). On the other hand in case of action of an earthquake, the dynamic characteristics of a soil have been referred traditionally to the speed of the shear waves that can be generated in a site, together with the characteristics of plasticity and damping of the soil. The Principle of the energy conservation explains why a shear upward propagating seismic wave can be amplified when travelling from a medium with high shear wave velocity (rock) to other medium with lower velocity (soil deposit), as it happened in the earthquake of Mexico of 1985. This amplification is a function of the speed gradient or of the contrast of impedances in the border of both types of mediums. A method is proposed in this paper for the design of containment walls in different soils, suffering to the action of an earthquake, based on the Performance-Based Seismic Design.

Performance profiles of football teams in the UEFA Champions League considering situational efficiency

Performance of football teams varies constantly due to the dynamic nature of this sport, whilst the typical performance and its spread can be represented by profiles combining different performance-related variables based on data from multiple matches. The current study aims to use a profiling technique to evaluate and compare match performance of football teams in the UEFA Champions League incorporating three situational variables (i.e. strength of team and opponent, match outcome and match location). Match statistics of 72 teams, 496 games across four seasons (2008-09 to 2012-13) of this competition were analysed. Sixteen performance-related events were included: shots, shots on target, shots from open play, shots from set piece, shots from counter attack, passes, pass accuracy (%), crosses, through balls, corners, dribbles, possession, aerial success (%), fouls, tackles, and yellow cards. Teams were classified into three levels of strength by a k-cluster analysis. Profiles of overall performance and profiles incorporating three situational variables for teams of all three levels of strength were set up by presenting the mean, standard deviation, median, lower and upper quartiles of the counts of each event to represent their typical performances and spreads. Means were compared by using one-way ANOVA and independent sample t test (for match location, home and away differences), and were plotted into the same radar charts after unifying all the event counts by standardised score. Established profiles can present straightforwardly typical performances of football teams of different levels playing in different situations, which could provide detailed references for coaches and analysts to evaluate performances of upcoming opposition and of their own.

Shake-table tests of a reinforced concrete frame designed following modern codes: seismic performance and damage evaluation

This paper presents shake-table tests conducted on a two-fifths-scale reinforced concrete frame representing a conventional construction design under current building code provisions in the Mediterranean area. The structure was subjected to a sequence of dynamic tests including free vibrations and four seismic simulations in which a historical ground motion record was scaled to levels of increasing intensity until collapse. Each seismic simulation was associated with a different level of seismic hazard, representing very frequent, frequent, rare and very rare earthquakes. The structure remained basically undamaged and within the inter-story drift limits of the "immediate occupancy" performance level for the very frequent and frequent earthquakes. For the rare earthquake, the specimen sustained significant damage with chord rotations of up to 28% of its ultimate capacity and approached the upper bound limit of inter-story drift associated with "life safety". The specimen collapsed at the beginning of the "very rare" seismic simulation. Besides summarizing the experimental program, this paper evaluates the damage quantitatively at the global and local levels in terms of chord rotation and other damage indexes, together with the energy dissipation demands for each level of seismic hazard. Further, the ratios of column-to-beam moment capacity recommended by Eurocode 8 and ACI-318 to guarantee the formation of a strong column-weak beam mechanism are examined.