An Introduction to MultiAgent Systems.

This paper presents control and coordination algorithms for groups of vehicles. The focus is on autonomous vehicle networks performing distributed sensing tasks where each vehicle plays the role of a mobile tunable sensor. The paper proposes gradient descent algorithms for a class of utility functions which encode optimal coverage and sensing policies. The resulting closed-loop behavior is adaptive, distributed, asynchronous, and verifiably correct.

Coverage control for mobile sensing networks

With its roots in kinetic theory and the cellular automaton concept, the lattice-Boltzmann (LB) equation can be used to obtain continuum flow quantities from simple and local update rules based on particle interactions. The simplicity of formulation and its versatility explain the rapid expansion of the LB method to applications in complex and multiscale flows. We review many significant developments over the past decade with specific examples. Some of the most active developments include the entropic LB method and the application of the LB method to turbulent flow, multiphase flow, and deformable particle and fiber suspensions. Hybrid methods based on the combination of the Eulerian lattice with a Lagrangian grid system for the simulation of moving deformable boundaries show promise for more efficient applications to a broader class of problems. We also discuss higherorder boundary conditions and the simulation of microchannel flow with finite Knudsen number. Additionally, the remarkable scalability of the LB method for parallel processing is shown with examples. Teraflop simulations with the LB method are routine, and there is no doubt that this method will be one of the first candidates for petaflop computational fluid dynamics in the near future.

Lattice-Boltzmann Method for Complex Flows

This chapter introduces the subject of statistical pattern recognition (SPR). It starts by considering how features are defined and emphasizes that the nearest neighbor algorithm achieves error rates comparable with those of an ideal Bayes’ classifier. The concepts of an optimal number of features, representativeness of the training data, and the need to avoid overfitting to the training data are stressed. The chapter shows that methods such as the support vector machine and artificial neural networks are subject to these same training limitations, although each has its advantages. For neural networks, the multilayer perceptron architecture and back-propagation algorithm are described. The chapter distinguishes between supervised and unsupervised learning, demonstrating the advantages of the latter and showing how methods such as clustering and principal components analysis fit into the SPR framework. The chapter also defines the receiver operating characteristic, which allows an optimum balance between false positives and false negatives to be achieved.

Statistical Pattern Recognition

Tumors generate physical forces during growth and progression. These physical forces are able to compress blood and lymphatic vessels, reducing perfusion rates and creating hypoxia. When exerted directly on cancer cells, they can increase cells' invasive and metastatic potential. Tumor vessels-while nourishing the tumor-are usually leaky and tortuous, which further decreases perfusion. Hypoperfusion and hypoxia contribute to immune evasion, promote malignant progression and metastasis, and reduce the efficacy of a number of therapies, including radiation. In parallel, vessel leakiness together with vessel compression causes a uniformly elevated interstitial fluid pressure that hinders delivery of blood-borne therapeutic agents, lowering the efficacy of chemo- and nanotherapies. In addition, shear stresses exerted by flowing blood and interstitial fluid modulate the behavior of cancer and a variety of host cells. Taming these physical forces can improve therapeutic outcomes in many cancers.

/pdf/the-role-of-mechanical-forces-in-tumor-growth-and-therapy-35zbud1t6d.pdf

The role of mechanical forces in tumor growth and therapy.

We describe here a rigorous and accurate model for the simulation of three-dimensional deformable particles (DPs). The method is very versatile, easily simulating various types of deformable particles such as vesicles, capsules, and biological cells. Each DP is resolved explicitly and advects within the surrounding Newtonian fluid. The DPs have a preferred rest shape (e.g., spherical for vesicles, or biconcave for red blood cells). The model uses a classic hybrid system: an Eulerian approach is used for the Navier-Stokes solver (the lattice Boltzmann method) and a Lagrangian approach for the evolution of the DP mesh. Coupling is accomplished through the lattice Boltzmann velocity field, which transmits force to the membranes of the DPs. The novelty of this method resides in its ability (by design) to simulate a large number of DPs within the bounds of current computational limitations: our simple and efficient approach is to (i) use the lattice Boltzmann method because of its acknowledged efficiency at low Reynolds number and its ease of parallelization, and (ii) model the DP dynamics using a coarse mesh (approximately 500 nodes) and a spring model constraining (if necessary) local area, total area, cell volume, local curvature, and local primary stresses. We show that this approach is comparable to the more common - yet numerically expensive - approach of membrane potential function, through a series of quantitative comparisons. To demonstrate the capabilities of the model, we simulate the flow of 200 densely packed red blood cells - a computationally challenging task. The model is very efficient, requiring of the order of minutes for a single DP in a 50 microm x 40 microm x 40 microm simulation domain and only hours for 200 DPs in 80 microm x 30 microm x 30 microm . Moreover, the model is highly scalable and efficient compared to other models of blood cells in flow, making it an ideal and unique tool for studying blood flow in microvessels or vesicle or capsule flow (or a mixture of different particles). In addition to directly predicting fluid dynamics in complex suspension in any geometry, the model allows determination of accurate, empirical rules which may improve existing macroscopic, continuum models.

/pdf/modeling-the-flow-of-dense-suspensions-of-deformable-yof8x55iwm.pdf

Modeling the flow of dense suspensions of deformable particles in three dimensions.

Emergencies in industrial warehouses are a major concern for fire-fighters. The large dimensions, together with the development of dense smoke that drastically reduces visibility, represent major c...

/pdf/a-robot-swarm-assisting-a-human-fire-fighter-29ctyjolu6.pdf

A Robot Swarm Assisting a Human Fire-Fighter

This paper presents robust methods for determining the order of a sequence of stripes captured in an uncoded structured light scanning system, i.e. where all the stripes are projected with uniform colour, width and spacing. A single bitmap image shows a pattern of vertical stripes from a projected source, which are deformed by the surface of the target object. If a correspondence can be determined between the projected stripes and those captured in the bitmap, a spatial measurement of the surface can be derived using standard rangefinding methods. Previous work has uniquely encoded each stripe, such as by colour or width, in order to avoid ambiguous stripe identification. However, colour coding suffers due to uneven colour reflection, and a variable width code reduces the measured resolution. To avoid these problems, we simplify the projection as a uniform stripe pattern, and devise novel methods for correctly indexing the stripes, including a new common inclination constraint and occlusion classification. We give definitions of patches and the continuity of stripes, and measure the success of these methods. Thus we eliminate the need for coding, and reduce the accuracy required of the projected pattern; and, by dealing with stripe continuity and occlusions in a new manner, provide general methods which have relevance to many structured light problems.

/pdf/methods-for-indexing-stripes-in-uncoded-structured-light-3z3kikreie.pdf

Methods for Indexing Stripes in Uncoded Structured Light Scanning Systems

This study aims to examine managers’ perceptions of digital banking’s (DB) effect on customer experience and banks’ financial performance.,The research uses interviews from the senior UK bank managers to gather their views on DB’s impact on customer experience and financial performance. The interviews were thematically analysed to produce results and a model.,The attributes affecting DB experience are as follows: service quality, functional quality, perceived value, service customisation, service speed, employee–customer engagement, brand trust, DB innovation, perceived usability and perceived risk. They affect customer experience, satisfaction and loyalty and financial performance. The research revealed relationships amongst these attributes (e.g. brand trust and loyalty).,The study is a UK bank specific and can be replicated in other developed countries’ banks, helping in further comparison. However, DB is conducted globally, which implies that the findings are robust enough to be potentially applied in other countries. The proposed model shows customer experience drivers and outcomes through managers’ views, which can be theoretically tested.,The findings suggest important attributes (as above) for consideration to improve DB customer experience and financial performance. They show the relevance of employee–customer interaction, service personalisation, value proposition, quality service offering and DB experience, which have useful implications for improving DB design and interactive marketing.,Gauging DB customer experience as perceived by bank managers has not been studied in this way, highlighting DB effectiveness, which is important for multi-channel marketing and banks’ financial performance, and advances theory.

/pdf/digital-banking-customer-experience-and-financial-1jfdtb88f6.pdf

Digital Banking, Customer Experience and Financial Performance: UK Bank Managers’ Perceptions

Mobile ad-hoc networks (MANETs) are communication networks that do not rely on fixed, preinstalled communication devices like base stations or predefined communication cells. MANETs are wireless networks consisting of mobile nodes which are characterized by their decentralized organization and the potentially high dynamics of the network structure. Therefore, MANETs are ideally suitable for applications with multi-robot systems. One of the most promising applications of a multi-robot system is to assist humans in urban search and rescue (USAR) scenarios in the aftermath of natural or man-made disasters. We are focusing on an ad-hoc network communication system with the mobile robots being communication nodes offering a robust communication infrastructure. Main disaster scenario covered by our system is a large industrial warehouse in fire, described in the GUARDIANS project funded by the European Union. In this scenario, black smoke may fill large space of the warehouse that makes it very difficult for the firefighters to orientate themselves in the building which in turn will usually limit
the action space of the firefighters. In order to increase the coverage area of the fire fighters the ad-hoc network has to provide position data to support localization of the mobile robots and humans, which might be of great importance to guide the humans and robots to specific targets and locations or to quickly exit the search area. In our proposed approach a cell-based network with master nodes in each cell forms the basic structure of the network. Some nodes formed by speciallyequipped robots act as beacons to uniformly span the network. These robots have a role as reference points when positioning other mobile robots or humans and at the same time form the infrastructure to support communication all over the
search area. A combination of distance and radio signal quality measurements as well as dedicated swarming behaviors of the robots are capable of maintaining suitable distribution of the robots even in the presence of walls that obstruct the radio signals. Communications standards considered for the ad-hoc network are Wireless LAN, Bluetooth and QigBee. All are integrated on a miniature robot for real experiments.
The features of the network are studied analytically, in simulations as well
as in experiments to verify the results. Furthermore, frequency and power managements are also taken into consideration to ensure robustness of communication in the network.

Lyuba Alboul

Papers

Modeling the flow of dense suspensions of deformable particles in three dimensions.

A Robot Swarm Assisting a Human Fire-Fighter

Methods for Indexing Stripes in Uncoded Structured Light Scanning Systems

Digital Banking, Customer Experience and Financial Performance: UK Bank Managers’ Perceptions

Ad-hoc Network Communication Infrastructure for Multi-robot Systems in Disaster Scenarios