Hybrid coordination of reinforcement learning-based behaviors for AUV control

This paper proposes a hybrid coordination method for behavior-based control architectures. The hybrid method takes advantages of the robustness and modularity in competitive approaches as well as optimized trajectories in cooperative ones. This paper shows the feasibility of applying this hybrid method with a 3D-navigation to an autonomous underwater vehicle (AUV). The behaviors are learnt online by means of reinforcement learning. A continuous Q-learning implemented with a feed-forward neural network is employed. Realistic simulations were carried out. The results obtained show the good performance of the hybrid method on behavior coordination as well as the convergence of the behaviors

Towards Direct Policy Search Reinforcement Learning for Robot Control

This paper proposes a high-level reinforcement learning (RL) control system for solving the action selection problem of an autonomous robot. Although the dominant approach, when using RL, has been to apply value function based algorithms, the system here detailed is characterized by the use of direct policy search methods. Rather than approximating a value function, these methodologies approximate a policy using an independent function approximator with its own parameters, trying to maximize the future expected reward. The policy based algorithm presented in this paper is used for learning the internal state/action mapping of a behavior. In this preliminary work, we demonstrate its feasibility with simulated experiments using the underwater robot GARBI in a target reaching task

Perspectives of auto-correcting lens distortions in mosaic-based underwater navigation

When unmanned underwater vehicles (UUVs) perform missions near the ocean floor, optical sensors can be used to improve local navigation. Video mosaics allow to efficiently process the images acquired by the vehicle, and also to obtain position estimates. We discuss in this paper the role of lens distortions in this context, proving that degenerate mosaics have their origin not only in the selected motion model or in registration errors, but also in the cumulative effect of radial distortion residuals. Additionally, we present results on the accuracy of different feature-based approaches for self-correction of lens distortions that may guide the choice of appropriate techniques for correcting distortions

Probabilistic sonar scan matching for an AUV

This paper proposes MSISpIC, a probabilistic sonar scan matching algorithm for the localization of an autonomous underwater vehicle (AUV). The technique uses range scans gathered with a Mechanical Scanning Imaging Sonar (MSIS), the robot displacement estimated through dead-reckoning using a Doppler velocity log (DVL) and a motion reference unit (MRU). The proposed method is an extension of the pIC algorithm. An extended Kalman filter (EKF) is used to estimate the robot-path during the scan in order to reference all the range and bearing measurements as well as their uncertainty to a scan fixed frame before registering. The major contribution consists of experimentally proving that probabilistic sonar scan matching techniques have the potential to improve the DVL-based navigation. The algorithm has been tested on an AUV guided along a 600 m path within an abandoned marina underwater environment with satisfactory results

5-Fluorouracil-loaded poly(ε-caprolactone) nanoparticles combined with phage E gene therapy as a new strategy against colon cancer

This work aimed to develop a new therapeutic approach to increase the efficacy of 5-fluorouracil (5-FU) in the treatment of advanced or recurrent colon cancer. 5-FU-loaded biodegradable poly(ε-caprolactone) nanoparticles (PCL NPs) were combined with the cytotoxic suicide gene E (combined therapy). The SW480 human cancer cell line was used to assay the combined therapeutic strategy. This cell line was established from a primary adenocarcinoma of the colon and is characterized by an intrinsically high resistance to apoptosis that correlates with its resistance to 5-FU. 5-FU was absorbed into the matrix of the PCL NPs during synthesis using the interfacial polymer disposition method. The antitumor activity of gene E from the phage ϕX174 was tested by generating a stable clone (SW480/12/E). In addition, the localization of E protein and its activity in mitochondria were analyzed. We found that the incorporation of 5-FU into PCL NPs (which show no cytotoxicity alone), significantly improved the drug's anticancer activity, reducing the proliferation rate of colon cancer cells by up to 40-fold when compared with the nonincorporated drug alone. Furthermore, E gene expression sensitized colon cancer cells to the cytotoxic action of the 5-FU-based nanomedicine. Our findings demonstrate that despite the inherent resistance of SW480 to apoptosis, E gene activity is mediated by an apoptotic phenomenon that includes modulation of caspase-9 and caspase-3 expression and intense mitochondrial damage. Finally, a strongly synergistic antiproliferative effect was observed in colon cancer cells when E gene expression was combined with the activity of the 5-FU-loaded PCL NPs, thereby indicating the potential therapeutic value of the combined therapy.

Adaptive intelligent agent approach to guide the web navigation on the PLAN-G distance learning platform

Hypermedia systems based on the Web for open distance education are becoming increasinglypopular as tools for user-driven access learning information. Adaptive hypermedia is a new direction in research within the area of user-adaptive systems, to increase its functionality by making it personalized [Eklu 961. This paper sketches a general agents architecture to include navigationaladaptability and user-friendly processes which would guide and accompany the student during hislher learning on the PLAN-G hypermedia system (New Generation Telematics Platform to Support Open and Distance Learning), with the aid of computer networks and specifically WWW technology [Marz 98-1] [Marz 98-2]. The PLAN-G actual prototype is successfully used with some informatics courses (the current version has no agents yet). The propased multi-agent system, contains two different types of adaptive autonomous software agents: Personal Digital Agents {Interface), to interacl directly with the student when necessary; and Information Agents (Intermediaries), to filtrate and discover information to learn and to adapt navigation space to a specific student

WMR navigation using local potential field corridors and narrow local occupancy grid perception

Path planning and control strategies applied to autonomous mobile robots should fulfil safety rules as well as achieve final goals. Trajectory planning applications should be fast and flexible to allow real time implementations as well as environment interactions. The methodology presented uses the on robot information as the meaningful data necessary to plan a narrow passage by using a corridor based on attraction potential fields that approaches the mobile robot to the final desired configuration. It employs local and dense occupancy grid perception to avoid collisions. The key goals of this research project are computational simplicity as well as the possibility of integrating this method with other methods reported by the research community. Another important aspect of this work consist in testing the proposed method by using a mobile robot with a perception system composed of a monocular camera and odometers placed on the two wheels of the differential driven motion system. Hence, visual data are used as a local horizon of perception in which trajectories without collisions are computed by satisfying final goal approaches and safety criteria

Mission control system for dam inspection with an AUV

This paper presents a complete control architecture that has been designed to fulfill predefined missions with an autonomous underwater vehicle (AUV). The control architecture has three levels of control: mission level, task level and vehicle level. The novelty of the work resides in the mission level, which is built with a Petri network that defines the sequence of tasks that are executed depending on the unpredictable situations that may occur. The task control system is composed of a set of active behaviours and a coordinator that selects the most appropriate vehicle action at each moment. The paper focuses on the design of the mission controller and its interaction with the task controller. Simulations, inspired on an industrial underwater inspection of a dam grate, show the effectiveness of the control architecture

ICTINEUAUV wins the First SAUC-E competition

A pioneer team of students of the University of Girona decided to design and develop an autonomous underwater vehicle (AUV) called ICTINEU-AUV to face the Student Autonomous Underwater Challenge-Europe (SAUC-E). The prototype has evolved from the initial computer aided design (CAD) model to become an operative AUV in the short period of seven months. The open frame and modular design principles together with the compatibility with other robots previously developed at the lab have provided the main design philosophy. Hence, at the robot's core, two networked computers give access to a wide set of sensors and actuators. The Gentoo/Linux distribution was chosen as the onboard operating system. A software architecture based on a set of distributed objects with soft real time capabilities was developed and a hybrid control architecture including mission control, a behavioural layer and a robust map-based localization algorithm made ICTINEU-AUV the winning entry

On AUV control architecture

This paper surveys control architectures proposed in the literature and describes a control architecture that is being developed for a semi-autonomous underwater vehicle for intervention missions (SAUVIM) at the University of Hawaii. Conceived as hybrid, this architecture has been organized in three layers: planning, control and execution. The mission is planned with a sequence of subgoals. Each subgoal has a related task supervisor responsible for arranging a set of pre-programmed task modules in order to achieve the subgoal. Task modules are the key concept of the architecture. They are the main building blocks and can be dynamically re-arranged by the task supervisor. In our architecture, deliberation takes place at the planning layer while reaction is dealt through the parallel execution of the task modules. Hence, the system presents both a hierarchical and an heterarchical decomposition, being able to show a predictable response while keeping rapid reactivity to the dynamic environment

El Bosc, element clau d'un paisatge i d'una comunitat

Aquest treball és un resum de la Memòria de Llicenciatura El Bosc, element clau d’un paisatge i d’una comunitat. La seva percepció per part dels pagesos de l’Alta Garrotxa. Assaig metodològic, llegida el 18 de desembre de 1981 a la Universitat Autònoma de Barcelona

Role of Notch signaling in the skin and cornea

Résumé : La voie de signalisation Notch est essentielle pour la différentiation de l'épiderme lors du développement embryonnaire de la peau. Il a été démontré que l'inactivation de Notch1 dans la peau de souris conduit à une hyperplasie de l'épiderme ainsi qu'à la formation subséquente de carcinomes basocellulaires ainsi que de plaques cornéennes. L'inactivation de Notch1 dans la cornée combinée à des lésions mécaniques démontre que les cellules progénitrices de la cornée se différentient en un épithélium hyperplasique et kératinisé comme la peau. Ce changement de destinée cellulaire conduit à une cécité cornéenne et implique des processus non-autonomes aux cellules épithéliales, caractérisés par la sécrétion de FGF-2 par l'épithélium Notch1-/- suivi d'une vascularisation et d'un remaniement du stroma sous-jacent. La déficience en vitamine A est connu comme cause de lésions cornéennes humaines (xérophtalmie sévère). En accord, nous avons trouvé que la signalisation Notch1 était liée au métabolisme de la vitamine A par la régulation de l'expression de CRBP1, nécessaire pour générer un pool de rétinol intracellulaire. La perte de Notch1 dans l'épiderme, l'autre récepteur de la famille présent dans la peau marine, ne conduit pas à un phénotype manifeste. Cependant, l'inactivation dans l'épiderme de Notch1 et Notch2 ensemble, ou de RBP-J, induit une dermatite atopique (DA) sévère chez les souris. De même, les patients souffrants de DA mais pas ceux souffrant de psoriasis ou de lichen plan, ont une réduction marquée de l'expression des récepteurs Notch dans la peau. La perte de Notch dans les keratinocytes conduit à une activation de la voie NF-κB, ce qui ensuite induit la production de TSLP, une cytokine profondément impliquée dans la pathogenèse de la DA. Nous démontrons génétiquement que TSLP est responsable de la DA ainsi que du développent d'un syndrome myéloprolifératif non-autonome aux cellules induit par le G-CSF. Cependant, ces souris avec une inactivation dans l'épiderme de Notch1 et Notch2 et aussi incapables de répondre au TSLP développent des tumeurs invasive sévères caractérisées par une haute activité de signalisation ß-catenin. TSLPR est identifié comme un potentiel suppresseur de tumeur non-autonome aux cellules tumorales; la transplantation de cellules hématopoïétiques TSLPR-/- dans des souris déficientes pour Notch est suffisant pour causer des tumeurs. Summary : The Notch pathway is essential for proper epidermal differentiation during embryonic skin development. It has previously been demonstrated that Notch1 inactivation in marine skin results in epidermal hyperplasia and subsequent formation of basal cell carcinoma-like (BCC-like) tumors as well as corneal plaques. Inducible ablation of Notch1 in the cornea combined with mechanical wounding show that Notch1 deficient corneal progenitor cells differentiate into a hyperplasic, keratinized, skin-like epithelium. This cell fate switch leads to corneal blindness and involves cell non-autonomous processes, characterized by secretion of FGF-2 through Notch1-/- epithelium followed by vascularisation and remodelling of the underlying stroma. Vitamin A deficiency is known to induce a similar corneal defect in humans (severe xerophthalmia). Accordingly, we found that Notch1 signaling is linked to vitamin A metabolism by regulating the expression of CRBP1, required to generate a pool of intracellular retinol. Epidermal loss of Notch2, the other Notch receptor present in marine skin, doesn't lead to any overt phenotypes. However, postnatal epidermis-specific inactivation of both Notch1 and Notch2, or of RBP-J, induces the development of a severe form of atopic dermatitis (AD) in mice. Likewise, patients suffering from AD, but not psoriasis or lichen planas, have a marked reduction of Notch receptor expression in the skin. Loss of Notch in keratinocytes leads to an activation of NF-κB signaling which in turn induces the production of Thymic stromal lymphopoietin (TSLP), a cytokine deeply implicated in the pathogenesis of AD. We genetically demonstrate that TSLP is responsible for AD as well as the development of a cell non-autonomous G-CSF induced myeloproliferative disorder (MPD) in mice. However, these mice with conditional epidermal inactivation of Notch1 and Notch2 as well as incapable to respond to TSLP develop severe invasive tumors characterized by high ß-catenin signaling activity. TSLPR is identified as a potential cell non-autonomous tumor suppressor; transplantation of TSLPR-/- hematopoietic cells into epidermal Notch deficient mice is sufficient to cause tumors.

Embodied Autonomy in Digital Ecosystems: From Bio-Inspired Agents to Cognitive Systems

This paper proposes to promote autonomy in digital ecosystems so that it provides agents with information to improve the behavior of the digital ecosystem in terms of stability. This work proposes that, in digital ecosystems, autonomous agents can provide fundamental services and information. The final goal is to run the ecosystem, generate novel conditions and let agents exploit them. A set of evaluation measures must be defined as well. We want to provide an outline of some global indicators, such as heterogeneity and diversity, and establish relationships between agent behavior and these global indicators to fully understand interactions between agents, and to understand the dependence and autonomy relations that emerge between the interacting agents. Individual variations, interaction dependencies, and environmental factors are determinants of autonomy that would be considered. The paper concludes with a discussion of situations when autonomy is a milestone

La competència «El treball col·laboratiu» : una oportunitat per a incorporar les TIC en la didàctica universitària. Descripció de l'experiència amb la plataforma ACME (UdG)

L'article és una reflexió sobre els requisits de formació dels professionals que demana la societat del coneixement. Un dels objectius més importants que ha de tenir la universitat en la societat del coneixement és la formació de professionals competents que tinguin prou eines intel·lectuals per a enfrontar-se a la incertesa de la informació, a la consciència que aquesta té una data de caducitat a curt termini i a l'ansietat que això provoca. Però, a més, també han de ser capaços de definir i crear les eines de treball amb què donaran sentit i eficàcia a aquest coneixement mudable i mutant. Per això, l'espai europeu d'ensenyament superior prioritza la competència transversal del treball col·laboratiu amb l'objectiu de promoure un aprenentatge autònom, compromès i adaptat a les noves necessitats de l'empresa del segle xxi. En aquest context, es presenta l'entorn teòric que fonamenta el treball desenvolupat a la plataforma informàtica ACME, que uneix el treball col·laboratiu i l'aprenentatge semipresencial o blended learning. Així mateix, es descriuen amb detall alguns exemples de wikis, paradigma del treball col·laboratiu, fets en assignatures impartides per la Universitat de Girona en l'espai virtual ACME

Use of palliative radiotherapy in brain and bone metastases (VARA II study).

INTRODUCTION Metastases are detected in 20% of patients with solid tumours at diagnosis and a further 30% after diagnosis. Radiation therapy (RT) has proven effective in bone (BM) and brain (BrM) metastases. The objective of this study was to analyze the variability of RT utilization rates in clinical practice and the accessibility to medical technology in our region. PATIENTS AND METHODS We reviewed the clinical records and RT treatment sheets of all patients undergoing RT for BM and/or BrM during 2007 in the 12 public hospitals in an autonomous region of Spain. Data were gathered on hospital type, patient type and RT treatment characteristics. Calculation of the rate of RT use was based on the cancer incidence and the number of RT treatments for BM, BrM and all cancer sites. RESULTS Out of the 9319 patients undergoing RT during 2007 for cancer at any site, 1242 (13.3%; inter-hospital range, 26.3%) received RT for BM (n = 744) or BrM (n = 498). These 1242 patients represented 79% of all RT treatments with palliative intent, and the most frequent primary tumours were in lung, breast, prostate or digestive system. No significant difference between BM and BrM groups were observed in: mean age (62 vs. 59 yrs, respectively); gender (approximately 64% male and 36% female in both); performance status (ECOG 0-1 in 70 vs. 71%); or mean distance from hospital (36 vs. 28.6 km) or time from consultation to RT treatment (13 vs. 14.3 days). RT regimens differed among hospitals and between patient groups: 10 × 300 cGy, 5 × 400 cGy and 1x800cGy were applied in 32, 27 and 25%, respectively, of BM patients, whereas 10 × 300cGy was used in 49% of BrM patients. CONCLUSIONS Palliative RT use in BM and BrM is high and close to the expected rate, unlike the global rate of RT application for all cancers in our setting. Differences in RT schedules among hospitals may reflect variability in clinical practice among the medical teams.