Abstract
This paper introduces a mathematical model (together with a relaxed version) and solution approaches for the multi-facility glass container production planning (MF-GCPP) problem. The glass container industry covers the production of glass packaging (bottle and jars), where a glass paste is continuously distributed to a set of parallel molding machines that shape the finished products. Each facility has a set of furnaces where the glass paste is produced in order to meet the demand. Furthermore, final product transfers between facilities are allowed to face demand. The objectives include meeting demand, minimizing inventory investment and transportation costs, as well as maximizing the utilization of the production facilities. A novel mixed integer programming formulation is introduced for MF-GCPP and solution approaches applying heuristics and meta-heuristics based on mathematical programming are developed. A multi-population genetic algorithm defines for each individual the partitions of the search space to be optimized by the MIP solver. A variant of the fix-and-optimize improvement heuristic is also introduced. The computational tests are carried on instances generated from real-world data provided by a glass container company. The results show that the proposed methods return competitive results for smaller instances, comparing to an exact solver method. In larger instances, the proposed methods are able to return high quality solutions.

Abstract
Large forest fires are notorious for their environmental and socio-economic impacts and are assigned a disproportionately high percentage of the fire management budget. This study addresses extremely large fires (ELF, C2500 ha) in Portugal (2003-2013). We analysed the effect of fire-suppression force variation on ELF duration, size and growth rate, versus the effect of the concomitant fire environment (namely fuel and weather) conditions. ELF occurred in highly flammable landscapes and typically were impelled by extreme fire weather conditions. Allocation of suppression resources (normalized per unit of burned area or perimeter length) was disparate among fires, suggesting inadequate incident management. Fire-suppression effort did not affect time to containment modelled by survival analysis. Regression tree analysis indicated ELF spread to be negatively affected by higher fire-suppression resourcing, less severe fire weather, lower time to containment and higher presence of <9-year-old fuels, by decreasing order of importance; regional variability was relevant. Fire environment-to-fire suppression ratios of influence were 3: 1 for fire size and 1: 1 for fire growth rate, respectively, explaining 76 and 60 % of the existing variability. Results highlight the opportunistic nature of large-fire containment. To minimize the area burned by ELF, management and operational improvements leading to faster containment are recommended, rather than higher fire-suppression resourcing; more effective identification and exploration of containment opportunities are preferable to the accumulation of suppression resources.

Abstract
In this paper we present a hybrid approach to integrative clustering based on the p-median problem with clients' preferences. We formulate the problem of simultaneous clustering of a set of objects, characterized by two sets of features, as a bi-level p-median model. An exact approach involving a branch-and-cut method combined with the simulated annealing algorithm is used, that allows one to find a two-source clustering. The proposed approach is compared with some well-known mathematical optimisation based clustering techniques applied to the NCI-60 tumour cell line anticancer drug screen dataset. The results obtained demonstrate the applicability of our approach to find competitive integrative clusterings. Copyright © by the paper's authors.

Abstract
This study proposes an optimization model to improve the robustness of an existing bus schedule. Robustness represents the ability of schedules to absorb deviations from the timetable and to prevent their propagation through the daily operations. The model developed proposes an optimal assignment of arrival times and distribution of slacks among Time Control Points of a bus line, in order to minimize delays and anticipations from schedule. This required the use of data collected through GPS devices installed in buses, informing the location of buses during their daily operation. The robustness of bus schedules was evaluated through the quantification of delays and anticipations of real observations of bus shifts by comparison with the timetable. The performance measures used to evaluate robustness are the average delay (or anticipation) of buses by comparison with the timetable, and the probability that a passenger that arrives on time according to the timetable will miss the bus or have to wait more than a specified threshold at a Time Control Point. We also compared the improvement of the schedule proposed by the optimization model with the original schedule. The results obtained in a real-world case study, corresponding to a bus line operating in Porto, showed that the model could return an improved schedule for all performance measures considered when compared with the original schedule.

Abstract
The United States, Brazil, and Argentina are responsible for 83% of world’s soybean production. Together, they respond to more than 80% of soybean grains and soybean meal exported and for more than 60% of soybean oil exportation. This paper studies the soybean trade of these three major exporters with the top ten commercial partners of each one in order to examine the main factors that influence this relationship. We follow a network analysis approach to evaluate the level of interdependence between exporters and importers. Our research studies the three main soybean products: grain, meal, and oil. The findings seem to indicate that countries prefer importing soybean grains to process inside their borders due to commodity prices and logistics costs.

Abstract
Contemporary urban transportation networks are facing challenges to address the growing needs of mobility, all the while improving their economic gains and environmental sustainability. Several studies demonstrate that competitive alternatives to individual private transport are able to address these challenges, such as public transportation services. Thus, the need for optimising their operational efficiency and offer user-centric service delivery arises, with a range of challenges related to the inherent complexity of urban transportation networks as well as the range of dynamic elements involved in such systems. An innovative approach to this problem leverages personal mobile devices in combination with collaborative exchange of information. In this study a system was developed to combine information provided by travellers with data from public transport operators. The result is a rich model of the transportation network that enables the distribution of information in a personalized way and in real-time: the Seamless Mobility solution. Large-scale and expensive infrastructures, such as existing ticketing systems, constitute a threat to such flexibility and traveller access to services. As a result, a distributed architecture was targeted with the goal of integrating personal mobile devices in the infrastructure, with benefits for travellers and transport operators. The proposed solution integrates a broad scope of challenges, including application of secure mobile payments methods, data aggregation from different components and distribution based on relevance techniques. With the implementation of this solution we expect to positively impact the way travellers and transport operators interact, and contribute towards mobility services that are more agile and adequate, taking into account that mobility patterns vary from person to person, seasonally, and even throughout a day. (C) 2016 The Authors. Published by Elsevier B.V.