Building a Real-Time Testing Platform for Unmanned Ground Vehicles with UDP Bridge.

Perception and vehicle control remain major challenges in the autonomous driving domain. To find a proper system configuration, thorough testing is needed. Recent advances in graphics and physics simulation allow researchers to build highly realistic simulations that can be used for testing in safety-critical domains and inaccessible environments. Despite the high complexity of urban environments, it is the non-urban areas that are more challenging. Nevertheless, the existing simulators focus mainly on urban driving. Therefore, in this work, we describe our approach to building a flexible real-time testing platform for unmanned ground vehicles for indoor and off-road environments. Our platform consists of our original simulator, robotic operating system (ROS), and a bridge between them. To enable compatibility and real-time communication with ROS, we generate data interchangeable with real-life readings and propose our original communication solution, UDP Bridge, that enables up to 9.5 times faster communication than the existing solution, ROS#. As a result, all of the autonomy algorithms can be run in real-time directly in ROS, which is how we obtained our experimental results. We provide detailed descriptions of the components used to build our integrated platform.

Adaptive Hurst-Sensitive Active Queue Management.

An Active Queue Management (AQM) mechanism, recommended by the Internet Engineering Task Force (IETF), increases the efficiency of network transmission. An example of this type of algorithm can be the Random Early Detection (RED) algorithm. The behavior of the RED algorithm strictly depends on the correct selection of its parameters. This selection may be performed automatically depending on the network conditions. The mechanisms that adjust their parameters to the network conditions are called the adaptive ones. The example can be the Adaptive RED (ARED) mechanism, which adjusts its parameters taking into consideration the traffic intensity. In our paper, we propose to use an additional traffic parameter to adjust the AQM parameters-degree of self-similarity-expressed using the Hurst parameter. In our study, we propose the modifications of the well-known AQM algorithms: ARED and fractional order PIαDß and the algorithms based on neural networks that are used to automatically adjust the AQM parameters using the traffic intensity and its degree of self-similarity. We use the Fluid Flow approximation and the discrete event simulation to evaluate the behavior of queues controlled by the proposed adaptive AQM mechanisms and compare the results with those obtained with their basic counterparts. In our experiments, we analyzed the average queue occupancies and packet delays in the communication node. The obtained results show that considering the degree of self-similarity of network traffic in the process of AQM parameters determination enabled us to decrease the average queue occupancy and the number of rejected packets, as well as to reduce the transmission latency.

LiDAR Point Cloud Generation for SLAM Algorithm Evaluation.

With the emerging interest in the autonomous driving level at 4 and 5 comes a necessity to provide accurate and versatile frameworks to evaluate the algorithms used in autonomous vehicles. There is a clear gap in the field of autonomous driving simulators. It covers testing and parameter tuning of a key component of autonomous driving systems, SLAM, frameworks targeting off-road and safety-critical environments. It also includes taking into consideration the non-idealistic nature of the real-life sensors, associated phenomena and measurement errors. We created a LiDAR simulator that delivers accurate 3D point clouds in real time. The point clouds are generated based on the sensor placement and the LiDAR type that can be set using configurable parameters. We evaluate our solution based on comparison of the results using an actual device, Velodyne VLP-16, on real-life tracks and the corresponding simulations. We measure the error values obtained using Google Cartographer SLAM algorithm and the distance between the simulated and real point clouds to verify their accuracy. The results show that our simulation (which incorporates measurement errors and the rolling shutter effect) produces data that can successfully imitate the real-life point clouds. Due to dedicated mechanisms, it is compatible with the Robotic Operating System (ROS) and can be used interchangeably with data from actual sensors, which enables easy testing, SLAM algorithm parameter tuning and deployment.

Supervised Learning of Neural Networks for Active Queue Management in the Internet.

The paper examines the AQM mechanism based on neural networks. The active queue management allows packets to be dropped from the router's queue before the buffer is full. The aim of the work is to use machine learning to create a model that copies the behavior of the AQM PIα mechanism. We create training samples taking into account the self-similarity of network traffic. The model uses fractional Gaussian noise as a source. The quantitative analysis is based on simulation. During the tests, we analyzed the length of the queue, the number of rejected packets and waiting times in the queues. The proposed mechanism shows the usefulness of the Active Queue Management mechanism based on Neural Networks.

Efficient Feature Selection for Static Analysis Vulnerability Prediction.

Common software vulnerabilities can result in severe security breaches, financial losses, and reputation deterioration and require research effort to improve software security. The acceleration of the software production cycle, limited testing resources, and the lack of security expertise among programmers require the identification of efficient software vulnerability predictors to highlight the system components on which testing should be focused. Although static code analyzers are often used to improve software quality together with machine learning and data mining for software vulnerability prediction, the work regarding the selection and evaluation of different types of relevant vulnerability features is still limited. Thus, in this paper, we examine features generated by SonarQube and CCCC tools, to identify those that can be used for software vulnerability prediction. We investigate the suitability of thirty-three different features to train thirteen distinct machine learning algorithms to design vulnerability predictors and identify the most relevant features that should be used for training. Our evaluation is based on a comprehensive feature selection process based on the correlation analysis of the features, together with four well-known feature selection techniques. Our experiments, using a large publicly available dataset, facilitate the evaluation and result in the identification of small, but efficient sets of features for software vulnerability prediction.

Diffusion Model of a Non-Integer Order PIγ Controller with TCP/UDP Streams.

In this article, a way to employ the diffusion approximation to model interplay between TCP and UDP flows is presented. In order to control traffic congestion, an environment of IP routers applying AQM (Active Queue Management) algorithms has been introduced. Furthermore, the impact of the fractional controller PIγ and its parameters on the transport protocols is investigated. The controller has been elaborated in accordance with the control theory. The TCP and UDP flows are transmitted simultaneously and are mutually independent. Only the TCP is controlled by the AQM algorithm. Our diffusion model allows a single TCP or UDP flow to start or end at any time, which distinguishes it from those previously described in the literature.

Long-Range Dependent Traffic Classification with Convolutional Neural Networks Based on Hurst Exponent Analysis.

The paper examines the ability of neural networks to classify Internet traffic data in terms of self-similarity expressed by the Hurst exponent. Fractional Gaussian noise is used for the generation of synthetic data for modeling the genuine ones. It is presented that the trained model is capable of classifying the synthetic data obtained from the Pareto distribution and the real traffic data. We present the results of training for different optimizers of the cost function and a different number of convolutional layers in the neural network.

Late-onset bloodstream infections of Very-Low-Birth-Weight infants: data from the Polish Neonatology Surveillance Network in 2009-2011.

BACKGROUND: Late-Onset Bloodstream Infections (LO-BSI) continue to be one of the most important complications associated with hospitalization of infants born with very low birth weight (VLBW). The aims of this study were to assess the epidemiology of LO-BSI together with the risk factors and the distribution of causative pathogens at six Polish neonatal intensive care units that participated in the Polish Neonatology Surveillance Network from January 1, 2009 to December 31, 2011. METHODS: The surveillance covered 1,695 infants whose birth weights were <1501 grams (VLBW) in whom LO-BSI was diagnosed >72 hours after delivery. Case LO-BSI patients were defined according to NeoKISS. RESULTS: Four hundred twenty seven episodes of LO-BSI were diagnosed with a frequency of 25.3% and an incidence density of 6.7/1000 patient-days (pds). Results of our multivariate analysis demonstrated that surgical procedures and lower gestational age were significantly associated with the risk of LO-BSI. Intravascular catheters were used in infants with LO-BSI significantly more frequently and/or for longer duration: Central venous cathters (CVC) (OR 1.29) and Peripheral venous catheters (PVC) (OR 2.8), as well as, the total duration of total parenteral nutrition (13 vs. 29 days; OR 1.81). Occurrence of LO-BSI was significantly associated with increased the length of mechanical ventilation (MV) (OR 2.65) or the continuous positive airway pressure (CPAP) (OR 2.51), as well as, the duration of antibiotic use (OR 2.98). The occurrence of more than one infection was observed frequently (OR 9.2) with VLBW with LO-BSI. Microorganisms isolated in infants with LO-BSI were dominated by Gram-positive cocci, and predominantly by coagulase-negative staphylococci (62.5%). CONCLUSIONS: Independent risk factor for LO-BSI in VLBV infants are: low gestational age and requirement for surgery. The incidence rates of LO-BSI especially CVC-BSI were higher in the Polish NICUs surveillance than those of other national networks, similar to the central- and peripheral utilization ratio.

Cost-effective filtering of unreliable proximity detection results based on BLE RSSI and IMU readings using smartphones.

Indoor environments are a major challenge in the domain of location-based services due to the inability to use GPS. Currently, Bluetooth Low Energy has been the most commonly used technology for such services due to its low cost, low power consumption, ubiquitous availability in smartphones and the dependence of the signal strength on the distance between devices. The article proposes a system that detects the proximity between static (anchors) and moving objects, evaluates the quality of this prediction and filters out the unreliable results based on custom metrics. We define three metrics: two matrics based on RSSI and Intertial Measurement Unit (IMU) readings and one joint metric. This way the filtering is based on both, the external information (RSSI) and the internal information (IMU). To process the IMU data, we use machine learning activity recognition models (we apply feature selection and compare three models and choose the best one-Gradient Boosted Decision Trees). The proposed system is flexible and can be easily customized. The great majority of operations can be conducted directly on smartphones. The solution is easy to implement, cost-efficient and can be deployed in real-life applications (MICE industry, museums, industry).

Finding the best hardware configuration for 2D SLAM in indoor environments via simulation based on Google Cartographer.

One of the most challenging topics in robotics is simultaneous localization and mapping (SLAM) in the indoor environments. Due to the fact that Global Navigation Satellite Systems cannot be successfully used in such environments, different data sources are used for this purpose, among others light detection and ranging (LiDARs ), which have advanced from numerous other technologies. Other embedded sensors can be used along with LiDARs to improve SLAM accuracy, e.g. the ones available in the Inertial Measurement Units and wheel odometry sensors. Evaluation of different SLAM algorithms and possible hardware configurations in real environments is time consuming and expensive. In our study, we evaluate the accuracy of mapping and localization (based on Absolute Trajectory Error and Relative Pose Error). Our use case is a robot used for room decontamination. The results for a small room show that for our robot the best hardware configuration consists of three LiDARs 2D, IMU and wheel odometry sensors. On the other hand, for long hallways, a configuration with one LiDAR 3D sensor and IMU works better and more stable. We also described a general approach together with tools and procedures that can be used to find the best sensor setup in simulation.

Does surgical site infection after Caesarean section in Polish hospitals reflect high-quality patient care or poor postdischarge surveillance? Results from a 3-year multicenter study.

BACKGROUND: Caesarean sections (CSs) are associated with a high infection risk. Surgical site infection (SSI) incidence is among the markers of effectiveness of infection prevention efforts. The aim of this study was to analyze risk factors for SSI, incidence, and microbiology in patients who underwent CS. METHODS: The study was conducted during 2013-2015 using active infection surveillance in 5 Polish hospitals according to the European Centre for Disease Prevention and Control surveillance network known as HAI-Net. For each procedure, the following data were registered: age, American Society of Anesthesiologists score, procedure time, elective or emergency procedure, use of perioperative antibiotic prophylaxis, microbiology, the treatment used, and other information. RESULTS: SSI incidence was 0.5% and significant differences were noted among hospitals (between 0.1% and 1.8%), for different American Society of Anesthesiologists scales (between 0.2% and 4.8%) and different values of standardized SSI risk index (between 0.0% and 0.8%). In 3.1% of procedures, with no antibiotic prophylaxis, SSI risk was significantly higher. Deep infections dominated: 61.5% with superficial infections in only approximately 30% of cases and 2.6% of infections were detected postdischarge without readmissions. CONCLUSIONS: Results showed high incidence of SSI in Poland without perioperative antibiotic prophylaxis, and secondly, ineffective surveillance according to CS status, considering outpatient obstetric care. Without postdischarge surveillance, it is not possible to recognize the epidemiologic situation, and further, to set priorities and needs when it comes to infection prophylaxis, especially because such low incidence may indicate no need for improvement in infection control.

Device-associated pneumonia of very low birth weight infants in Polish Neonatal Intensive Care Units.

PURPOSE: Late-Onset Pneumonia (LO-PNEU) is still the most important complication associated with the hospitalization of infants with very low birth weight (<1501g). The purpose of this paper is to summarize the results of an ongoing surveillance program defining LO-PNEU as associated or not associated with respiratory support in the NICU and distribution of causative pathogens from the Polish Neonatology Surveillance Network (PNSN). MATERIALS AND METHODS: Surveillance of infections was conducted in the years 2009-2011 at six Polish NICUs. RESULTS: The incidence was 3.1/1000 NICU patient days (pds). The mean gestational age and birth weight among infants with LO-PNEU were significantly lower. The VAP incidence was of 18.2/1000 NICU pds for mechanically ventilated (MV) infants, while the rates for those receiving only CPAP were as low as 7.7/1000 NICU pds. MV significantly increased the risk of PNEU, but MV or CPAP for <10 days did not increase the risk of LO-PNEU. Significantly associated with LO-PNEU was the use of central or peripheral venous catheters and total parenteral nutrition for longer periods. Microorganisms isolated in cases of LO-PNEU were Gram-positive cocci (53.5%) and Gram-negative rods, with predominating E. coli. Non fermentative bacilli were significantly more frequent in cases of VAP than in other cases. CONCLUSIONS: Observed incidence rates associated with VAP and CPAP-PNEU, were higher than in other national surveillance systems and expressing the feasibility of lowering the risk of LO-PNEU and increasing patient safety. The incidence of pneumonia was found to be lower when using CPAP as compared to using MV.

Necrotising enterocolitis in preterm infants: epidemiology and antibiotic consumption in the Polish neonatology network neonatal intensive care units in 2009.

UNLABELLED: The aim of this study was to describe the epidemiology of necrotising enterocolitis (NEC), antibiotic consumption and the usefulness of microbiological tests in very low birth weight (VLBW) Polish newborns. METHODS: Prospective surveillance was performed in the year 2009 by local infection control teams. The study covered 910 infants hospitalized in six Polish neonatal intensive care units. Two kinds of indicators were used for the description of antibiotic usage: the duration of treatment (days of treatment, DOTs) and the defined daily dose (DDD). RESULTS: NEC incidence was 8.7% and fatality rate was 19%. Chorioamnionitis, late gestational age and low birth weight were identified as risk factors for NEC. Catheterization, mechanical ventilation and other selected procedures were used considerably longer in newborns with NEC than in the remaining neonates. Total usage of antibiotics reached 2.9 DDDs or 1.437 days; the average use of drugs per case of NEC amounted to 0.47 DDD or 23.2 DOTs. The level of antibiotic usage was analysed with correlation to microbiological tests performed and it was non-significantly greater in the group of children with NEC in whom the tests were performed. CONCLUSIONS: A high risk of developing NEC is closely associated with VLBW and with inflammation of the amnion during labour. We observed no relationship between the consumption of antibiotics in neonates with NEC and positive results of microbiological testing indicating sepsis accompanying NEC or gut colonization with pathogens.

Infections reported in newborns with very low birth weight who required surgical treatment. Data from the Polish Neonatology Surveillance Network.

AIM: To determine the risk of various forms of infections appearing in very low birth weight newborns (VLBW) during the period of 30 days after surgical treatment in hospitals which have perinatal care departments with neonatal intensive care units that form the Polish Neonatology Surveillance Network (PNSN). MATERIAL AND METHODS: Continuous prospective monitoring of infections was carried out from January 1st to December 31st 2009 in six neonatal intensive care units which form the Polish Neonatal Surveillance Network. (PNSN). 910 newborns with very low birth weight (VLBW) were included in the study programme. 91 (10%) of this group underwent 118 surgical interventions. 12 newborns needed two or more surgeries. The most common procedure was the closure of persistent ductus artesiosus (PDA) and photocoagulation of vascular damage in the eye fundus. RESULTS: In the period of 30 days after surgery the following were diagnosed: in 3 newborns - necrotizing enterocolitis (NEC), in 22 newborns - sepsis (BSI) and in 54 newborns - pneumonia (PNEU). Symptoms of BSI and PNEU were on average observed on the 10th day after surgical intervention, while in the case of NEC on the 17th day. The highest incidence of infection (148.4%) was observed after PDA closure and in connection with introducing a drain into the pleural cavity through the intercostal space. The incidence of PN EU (37.3%) was twice as high as the incidence of BSI (18.6%). Surgical procedure was a factor significantly increasing the risk of infection and morbidity (RR 2.1, P<001) In our investigations there was no case of the local infection of a surgical site. 11 newborns died (mortality was 12.1%). The most common bacterial strains found in our investigation were coagulase-negative Staphylococcus and Escherichia Coli. CONCLUSIONS: Taking into consideration the fact that surgical procedure in VLBW-newborns significantly increases the risk of pneumonia and to a minor degree the risk of NEC and BSI, further detailed investigation in the field of perisurgical epidemiology is needed in order to elaborate a more deeply oriented and more effective strategy of infection prevention in the group of VLBW newborns.

[Epidemiology of infections in very low birth weight infants. Polish Neonatology Network research]. / Epidemiologia Zakazen Noworodkówz Bardzo Mala Masa Urodzeniowa.badania Polskiej Sieci Neonatologicznej.

AIM OF STUDY: 1. prospective record of infections; 2. prevention of nosocomial infections by providing current data about infections, which are significant for making therapeutic decisions. MATERIAL AND METHODS: Recorded infections: early-onset sepsis (congenital), late-onset sepsis (acquired in hospital), necrotising enterocolitis (NEC), pneumonia. Infections were diagnosed and qualified on the basis of definitions of infections based on the National Nosocomial Infections Surveillance (NNIS) criteria, developed by CDC, USA, including modifications of German Neo-KISS programme. Infection control was realised as a part of common research project of "Polish Neonatology Network", appointed by the decision of the Minister of Science and Higher Education no. 669/E-215/BWSN- 0180/2008 dated 20.05.2008 r. The study was conducted by 6 Polish neonatology units, Microbiology Chair of Jagiellonian University Collegium Medicum and Institute of Theoretical and Applied Computer Science. Infants with birth weight lower than 1500 g were qualified for the study. RESULTS: Between 1.01.2009 and 31.12.2009, 910 patients were registered, i.e. 19.1% of VLBW infants born in that period. The conducted analysis showed significant differences between centres in gestational age, birth weight, hospitalisation, use of invasive procedures, antibiotics and parenteral feeding. Cumulative morbidity rate was 68.5% total. Blood infection (sepsis) was the most commonly observed type of infection: 268 cases - 43.1% of all recorded forms of infection. Pneumonia was diagnosed in 242 cases, 38.8% total. NEC constituted 12.7% studied infections (79 cases). Apart from NEC, the risk of other forms of infection differed between centres. Dominant etiologic factor of all infections were Gram-positive cocci, which constituted 565 isolated microorganisms. Among them coagulase-negative staphylococci (CNS) were the most common (41.7%), while Staphylococcus aureus was fourth most frequent etiologic factor of infections (6.3% total). CONCLUSIONS: 1. Introduction of unified definition and criteria for diagnosing infections and the use of morbidity rates enables comparative analysis of epidemiology of infections in neonatal intensive care units. 2. Due to significant differences observed between prophylactic and therapeutic procedures in various units, it is essential to propose Polish recommendations regarding control over etiology of infections and use of invasive procedures, such as intravenous catheters and ventilation. 3. It is essential to undertake action leading to fully rational antibiotic therapy, because overuse of antibiotics leads to bacterial resistance and increases incidence of infections.

Enterobacteriaceae infections of very low birth weight infants in Polish neonatal intensive care units: resistance and cross-transmission.

BACKGROUND: The aims of our study were analysis of the occurrence of infections by members of the Enterobacteriaceae family in 6 Polish neonatal intensive care units in 2009, their drug resistance, the epidemiology of extended-spectrum ß-lactamase (ESBL)-producing strains and the possibility of using modern tools of microbiology diagnosis in infection control, especially for the reduction of antimicrobial resistance. METHODS: A prospective surveillance covered 910 newborns. Case patients were defined as neonates with very low birth weight who had clinical signs of septicemia, pneumonia or necrotizing enterocolitis. Early-onset infection was defined as infection diagnosed within 3 days after delivery. RESULTS: The incidence of Enterobacteriaceae infections was 2.6/1000 patient-days. The risk of Enterobacteriaceae pneumonia increased with the length of hospitalization (P = 0.0356). The most common pathogen was Escherichia coli (12.4% of all strains, in early-onset infection 18.5%) and Klebsiella spp. (9.1% of all). The ESBL phenotype was found in 37% of isolates, of which 89.3% were producing CTX-M-type, 70.2% TEM-type and 8.5% SHV-type. Epidemic clones were detected in the 2 studied neonatal intensive care units: 6 of the 9 ESBL-positive Enterobacter cloacae and 16 of the 18 ESBL-positive Klebsiella pneumoniae strains were classified into 1 epidemic clone, which showed resistance to penicillin without inhibitors, amoxycillin/clavulanic acid, cephalosporins, aztreoname, aminoglycosides and trimethoprim/sulfamethoxazole. CONCLUSIONS: Enterobacteriaceae bacilli are a significant problem in neonatal intensive care units, especially in early-onset infection and for long hospitalized very low birth weight infants. The observed high drug resistance was in large part related to the dominance of epidemic strains as a result of horizontal transmission. The best way to reduce drug resistance would be adequate procedures of isolation and hand hygiene.

Early-onset infections of very-low-birth-weight infants in Polish neonatal intensive care units.

AIM: The objective of this study was to investigate the incidence, causes, the risk factors, etiologic agents and the outcomes of early-onset infections (EOIs) in very-low-birth-weight newborns in Polish neonatal intensive care units. METHODS: Continuous prospective infection surveillance conducted during 2009 at 6 Polish neonatal intensive care units and included 910 newborns whose birth weight was lower than 1500 g. Infections were defined according to the Gastmeier's criteria. EOIs were diagnosed <3 days after delivery. RESULTS: The frequency of early-onset septicemia (EOS) was 7.0% and of early-onset pneumonia (EO-pneumonia) 8.6%. The factors significantly increasing the risk of EOS were low gestational age, small birth weight, low score in the Clinical Risk Index for Babies and Apgar score as well as maternal chorioamnionitis. The perinatal prophylaxis did not have an influence on the occurrence of EOS. The factors considerably increasing the risk of EO-pneumonia were low scores in the Clinical Risk Index for Babies and Apgar scores, a low gestational age and bacterial vaginosis in the child's mother during pregnancy. The most important etiologic organisms were Gram-positive cocci (39.7% of all the infections, 47.8% in EOS), Streptococcus agalactiae (20% of the EOS), Gram-negative bacilli (33.3% isolates), yeast-like fungi (isolated in 7.9% of cases) and atypical bacteria (22% of the cases of EO-pneumonia). CONCLUSIONS: The observed frequency of EOS did not differ from the one described in the literature, whereas the frequency of EO-pneumonia was higher. The bacterial etiologies suggest the vertical transmission of the pathogens and a close relationship between the observed EOIs with maternal environment. The applied perinatal antibiotic prophylaxis was ineffective.