An experimental characterization of a tunable separation device

By the use of an experimental setup for microfluidic flows, we have characterized the separation and concentration characteristics of the so-called Trilobite^? separation unit. Our separation unit consists of microfluidic channels and an elliptical separation geometry with a solid and a permeable wall region. We show that it is possible to adjust the thickness of different flow layers by changing the flow rates and pressure drop over the permeable wall. For high pressure drops, the separator shows promising concentration characteristics. For low pressure drops, an increase in flow rate results in a thinning of the flow layers. For sufficiently high flow rates, it should therefore be possible to create flow layers sufficiently thin that the particle separation is entirely dominated by hydrodynamic forces. This, in turn, will enable clog-free particle separation.     

Creating geometrically robust designs for highly sensitive problems using topology optimization

Resonance and wave-propagation problems are known to be highly sensitive towards parameter variations. This paper discusses topology optimization formulations for creating designs that perform robustly under spatial variations for acoustic cavity problems. For several structural problems, robust topology optimization methods have already proven their worth. However, it is shown that direct application of such methods is not suitable for the acoustic problem under consideration. A new double filter approach is suggested which makes robust optimization for spatial variations possible. Its effect and limitations are discussed. In addition, a known explicit penalization approach is considered for comparison. For near-uniform spatial variations it is shown that highly robust designs can be obtained using the double filter approach. It is finally demonstrated that taking non-uniform variations into account further improves the robustness of the designs.     

A robot learning from demonstration framework to perform force-based manipulation tasks

This paper proposes an end-to-end learning from demonstration framework for teaching force-based manipulation tasks to robots. The strengths of this work are manyfold. First, we deal with the problem of learning through force perceptions exclusively. Second, we propose to exploit haptic feedback both as a means for improving teacher demonstrations and as a human–robot interaction tool, establishing a bidirectional communication channel between the teacher and the robot, in contrast to the works using kinesthetic teaching. Third, we address the well-known what to imitate? problem from a different point of view, based on the mutual information between perceptions and actions. Lastly, the teacher’s demonstrations are encoded using a Hidden Markov Model, and the robot execution phase is developed by implementing a modified version of Gaussian Mixture Regression that uses implicit temporal information from the probabilistic model, needed when tackling tasks with ambiguous perceptions. Experimental results show that the robot is able to learn and reproduce two different manipulation tasks, with a performance comparable to the teacher’s one.     

Efficiently adapting graphical models for selectivity estimation

Query optimizers rely on statistical models that succinctly describe the underlying data. Models are used to derive cardinality estimates for intermediate relations, which in turn guide the optimizer to choose the best query execution plan. The quality of the resulting plan is highly dependent on the accuracy of the statistical model that represents the data. It is well known that small errors in the model estimates propagate exponentially through joins, and may result in the choice of a highly sub-optimal query execution plan. Most commercial query optimizers make the attribute value independence assumption: all attributes are assumed to be statistically independent. This reduces the statistical model of the data to a collection of one-dimensional synopses (typically in the form of histograms), and it permits the optimizer to estimate the selectivity of a predicate conjunction as the product of the selectivities of the constituent predicates. However, this independence assumption is more often than not wrong, and is considered to be the most common cause of sub-optimal query execution plans chosen by modern query optimizers. We take a step towards a principled and practical approach to performing cardinality estimation without making the independence assumption. By carefully using concepts from the field of graphical models, we are able to factor the joint probability distribution over all the attributes in the database into small, usually two-dimensional distributions, without a significant loss in estimation accuracy. We show how to efficiently construct such a graphical model from the database using only two-way join queries, and we show how to perform selectivity estimation in a highly efficient manner. We integrate our algorithms into the PostgreSQL DBMS. Experimental results indicate that estimation errors can be greatly reduced, leading to orders of magnitude more efficient query execution plans in many cases. Optimization time is kept in the range of tens of milliseconds, making this a practical approach for industrial-strength query optimizers.     

Combining the triangle method with thermal inertia to estimate regional evapotranspiration — Applied to MSG-SEVIRI data in the Senegal River basin

Spatially distributed estimates of evaporative fraction and actual evapotranspiration are pursued using a simple remote sensing technique based on a remotely sensed vegetation index (NDVI) and diurnal changes in land surface temperature. The technique, known as the triangle method, is improved by utilizing the high temporal resolution of the geostationary MSG-SEVIRI sensor. With 15 min acquisition intervals, the MSG-SEVIRI data allow for a precise estimation of the morning rise in land surface temperature which is a strong proxy for total daytime sensible heat fluxes. Combining the diurnal change in surface temperature, dT_s with an interpretation of the triangular shaped dT_s − NDVI space allows for a direct estimation of evaporative fraction. The mean daytime energy available for evapotranspiration (R_n − G) is estimated using several remote sensors and limited ancillary data. Finally regional estimates of actual evapotranspiration are made by combining evaporative fraction and available energy estimates. The estimated evaporative fraction (EF) and actual evapotranspiration (ET) for the Senegal River basin have been validated against field observations for the rainy season 2005. The validation results showed low biases and RMSE and R² of 0.13 [−] and 0.63 for EF and RMSE of 41.45 W m^− 2 and R² of 0.66 for ET.     

A Web services-based framework for building componentized digital libraries

We present a new Web services-based framework for building componentized digital libraries (DLs). We particularly demonstrate how traditional RDBMS technology can be easily deployed to support several common digital library services. Configuration and customization of the framework to build specialized systems is supported by a wizard-like tool which is based on a generic metamodel for DLs. Such a tool implements a workflow process that segments the DL design tasks into well-defined steps and drives the designer along these steps. Both the framework and the configuration tool are evaluated in terms of several performance and usability criteria. Our experimental evaluation demonstrates the feasibility and superior performance of our framework, as well as the effectiveness of the wizard tool for setting up DLs.     

Development of a genetic algorithm for component placement sequence optimization in printed circuit board assembly

This paper describes the development and evaluation of a custom application exploring the use of genetic algorithms (GA) to solve a component placement sequencing problem for printed circuit board (PCB) assembly. In the assembly of PCB’s, the component placement process is often the bottleneck, and the equipment to complete component placement is often the largest capital investment. The number of components placed on a PCB can range from few to hundreds. As a result, developing an application to determine an optimal or near-optimal placement sequence can translate into reduced cycle times for the overall assembly process and reduced assembly costs. A custom application was developed to evaluate the effectiveness of using GA’s to solve the component placement sequencing problem. A designed experiment was used to determine the best representation and crossover type, crossover rate, and mutation rate to use in solving a component sequencing problem for a PCB consisting of 10 components being placed on a single-headed placement machine. Three different representations (path, ordinal, and adjacency) and six appropriate crossover types (partially mapped, ordered, cycle, classical, alternating edges, and heuristic) were evaluated at three different mutation rates and at 11 crossover rates. Two algorithm response variables, the total distance traveled by the placement head and the algorithm solution efficiency (measured as number of generations and algorithm solution time) were used to evaluate the different GA applications. The combination of representation and crossover type along with mutation rate were found to be the most significant parameters in the algorithm design. In particular, path representation with order crossover was found to produce the best solution as measured by the total distance traveled as well as the solution generation efficiency. Increasing the mutation rate led to slightly improved solutions in terms of head travel, but also resulted in increased solution time.     

GMAC: An overlay multicast network for mobile agent platforms

The lack of proper support for multicast services in the Internet has hindered the widespread use of applications that rely on group communication services such as mobile software agents. Although they do not require high bandwidth or heavy traffic, these types of applications need to cooperate in a scalable, fair and decentralized way. This paper presents GMAC, an overlay network that implements all multicast related functionality–including membership management and packet forwarding–in the end systems. GMAC introduces a new approach for providing multicast services for mobile agent platforms in a decentralized way, where group members cooperate in a fair way, minimize the protocol overhead, thus achieving great scalability. Simulations comparing GMAC with other approaches, in aspects such as end-to-end group propagation delay, group latency, group bandwidth, protocol overhead, resource utilization and failure recovery, show that GMAC is a scalable and robust solution to provide multicast services in a decentralized way to mobile software agent platforms with requirements similar to MoviLog.