Abstract
This paper describes a group of robots for cleaning a simulated environment and proposes an efficient algorithm for navigation based on Pathfinding A *. No need for vision sensors. As a result it was observed that the robots can work cooperatively to clear the ground and that the navigation algorithm is effective in cleaning. In order to test its efficiency it was compared the combination of the Pathfinding A* algorithm and the decision algorithm proposed in this paper with Pathfinding A* and Euclidean distance, resulted in an improvement in time and distance traveled. © CLAWAR Association.

Abstract
This paper studies the use of deep convolutional neural networks to segment heart sounds into their main components. The proposed methods are based on the adoption of a deep convolutional neural network architecture, which is inspired by similar approaches used for image segmentation. Different temporal modeling schemes are applied to the output of the proposed neural network, which induce the output state sequence to be consistent with the natural sequence of states within a heart sound signal (S1, systole, S2, diastole). In particular, convolutional neural networks are used in conjunction with underlying hidden Markov models and hidden semi-Markov models to infer emission distributions. The proposed approaches are tested on heart sound signals from the publicly available PhysioNet dataset, and they are shown to outperform current state-of-the-art segmentation methods by achieving an average sensitivity of 93.9 and an average positive predictive value of 94 in detecting S1 and S2 sounds.

Abstract
Diesel-fueled buses have been replaced by Compressed Natural Gas (CNG) to minimize the high level of emissions in urban areas. However, differences in indoor exposure levels to Benzene, Toluene, Ethylbenzene and Xylene (BTEX) in those vehicles have not been investigated so far. The primary aim of this study was to determine if passengers are exposed to different BTEX levels when using buses powered by CNG or by diesel, and further explore if indoor levels are influenced by external air quality. For this purpose, BTEX air concentrations were measured in bus cabins (CNG and diesel), parking stations and in a background urban area using passive air samplers. Results showed that BTEX concentrations inside vehicles were higher than outside, but no significant differences were found between buses powered by CNG or by diesel. In CNG vehicles, high and significant positive correlation was found between benzene and the number of journeys in the same route (r(s) = 0.786, p < 0.05), vehicle operating time (r(s) = 0.738, p < 0.05), exposure time (r(s) = 0.714, p < 0.05) and exposure index (r(s) = 0.738, p < 0.05), but this was not observed for diesel vehicles. Benzene in bus cabins was found to be significantly below reference value for human health protection. However, excepting p-xylene, all other aromatic pollutants have a mean concentration significantly above the lowest effect level (p <= 0.002 for all comparisons). Additionally, higher BTEX levels in cabin buses than in outdoor air suggest the presence of other emission sources in indoor cabins. These findings emphasize the need for further studies to fully characterize indoor emission sources in order to minimize the negative impact of BTEX exposure to human health.

Abstract
This paper describes a solution method that was created with the objective of obtaining a more efficient finished goods distribution process for a food industry company. The finished goods distribution process involves the use of the companys own fleet to serve a specific group of customers, and the use of outsourcing transportation services that can make direct and transshipment customer deliveries. The complexity of the problem is due to the need to decide which customers should be served by each of the outsourcing transportation services, direct or transshipment, and to find cost efficient solutions for the multiple vehicle routing problems created. First, an original clustering method consisting of a logical division of the customer orders using a delivery ratio based on the transportation unit cost, distance and order weight, is used to define customer clusters by service type. Then, an exact method based on a mixed integer programming model, is used to obtain optimal vehicle routing solutions, for each cluster created. The solution method for the company real instances, proved able to reach the initial proposed objectives and obtain promising results that suggest an average reduction of 34% for the operational costs, when compared to the current distribution model of the company. © 2019, Springer Nature Switzerland AG.

Abstract
A reflective fiber optic sensor based on multimode interference for the measurement of relative humidity (RH) is proposed and experimentally demonstrated. The proposed probe is fabricated by fusion-splicing, approximately 30 mm long coreless fiber section to a single mode fiber. A hydrophilic agarose gel is coated on the coreless fiber, using the dip coating technique. When the incident light comes from the SMF to the CSF, the high-order modes are excited and propagate within the CSF. These excited modes interfere with one another as they propagate along whole CSF length, giving rise to a multimode interference (MMI). Since the effective refractive index of the agarose gel changes with the ambient relative humidity, as the environmental refractive index changes, the propagation constants for each guided mode within the CSF will change too, which leads to shifts in the output spectra. The proposed sensor has a great potential in real time RH monitoring, exhibiting a large range of operation with good stability. For RH variations in the range between 60 %RH and 98.5 %RH, the sensor presents a maximum sensitivity of 44.2 pm/%RH, and taking in consideration the interrogation system, a resolution of 1.1%RH is acquired. This sensor can be of interest for applications where a control of high levels of relative humidity is required.

Abstract
In an increasingly and globally competitive industry like the automotive sector, the continuous improvement of processes assumes a key role in the enhancement of effectiveness and efficiency. In line with this philosophy, the study undertaken of the work method and its inherent activities has pointed to time measurement as a gemba support tool of great potential in the optimization of the production process and the elimination of mudas. The present case study relates to the improvement of productivity of an assembly line dedicated to the manufacture of brake cables for the automotive industry. The application of the A3 methodology to this improvement project aims to monitor the evolution of key indicators as productivity per hour and the Overall Equipment Effectiveness (OEE) of the assembly line, to define the different improvement actions to be executed, and to achieve the defined target, and to validate the results of its implementation. The measurement of the different activities of the assembly line allowed to identify the major wastes (of material movements, operator movements, among others.) and to identify the most critical workstations that contribute for the unbalancing of the assembly line. The solutions implemented allowed to increase the productivity by 49% and, as a consequence, reduce the cycle time in 33%. The rearrangement and improvement of operations allowed also to increase the efficiency of the assembly line balancing in 11%. As the A3 methodology assumed a key role for this project, allowing the monitoring of the effectiveness of the different improvement actions implemented, it was standardized so that it could be applied to other improvement projects.