•The bottling problem in the wine industry is identified and formulated as a MIP.•Due to the high computational times of the MIP approach we provide a greedy heuristic.•For real size industrial ...instances our heuristic finds solutions in seconds.•This technique is a promising alternative for experience based scheduling methods.
In this work, we address the bottling scheduling problem that arises in the wine industry when the packing requests from clients need to be allocated to the production lines. This problem also appears in a large variety of industries, but especially in packaged food companies. Based on the operations of a large Chilean winery we worked with, we developed a MIP model that exhibits industry-specific features such as different types of wine resources and oenological process constraints. This model can be reduced to an n job, m parallel machine scheduling problem, which is known to be NP-hard, so we developed a greedy heuristic algorithm in order to find a feasible bottling schedule in a reduced computing time. We show that the proposed solution approach is a very promising alternative to efficient MIP solvers like CPLEX. Particularly, the greedy heuristic is able to schedule all the jobs in 98% of the test instances and the computational times are very reasonable even for large industrial cases.
•We propose a horizontal collaborative approach for the wine bottling scheduling problem.•We use a cooperative game approach where the characteristic function is computed heuristically.•We propose a ...maximum entropy methodology which simultaneously solves the partner selection and cost allocation problems.•Numerical experiments reveal that collaboration can have important positive effects.
This paper proposes a horizontal collaborative approach for the wine bottling scheduling problem. The opportunities for collaboration in this problem are due to the fact that many local wine producers are usually located around the same region and that bottling is a standard process. Collaboration among wineries is modeled as a cooperative game, whose characteristic function is derived from a mixed integer linear programming model. Real world instances of the problem are, however, unlikely to be solved to optimality due to its complex combinatorial structure and large dimension. This motivates the introduction of an approximated version of the original game, where the characteristic function is computed through a heuristic procedure. Unlike the exact game, the approximated game may violate the subadditivity property. Therefore, it turns relevant not only to find a stable cost allocation but also to find a coalition structure for selecting the best partition of the set of firms. We propose a maximum entropy methodology which can address these two problems simultaneously. Numerical experiments illustrate how this approach applies, and reveal that collaboration can have important positive effects in wine bottling scheduling decreasing delay by 33.4 to 56.9% when improvement heuristic solutions are used. In contrast to the exact game in which the grand coalition is always the best outcome, in the approximated game companies may be better forming smaller coalitions. We also devise a simple procedure to repair the characteristic function of the approximated game so that it recovers the subadditivity property.
•We propose a new image-inspired architecture for predicting crashes.•We analyze its predictive power by applying different oversampling methodologies.•We find that DCGAN with random undersampling ...maximize the predictive power.•We build a concatenated CNN of multiple inputs and compare with other methodologies.•By including the image-data, the predictive performance improves considerably.
In road safety, real-time crash prediction may play a crucial role in preventing such traffic events. However, much of the research in this line generally uses data aggregated every five or ten minutes. This article proposes a new image-inspired data architecture capable of capturing the microscopic scene of vehicular behavior. In order to achieve this, an accident-prediction model is built for a section of the Autopista Central urban highway in Santiago, Chile, based on the concatenation of multiple-input Convolutional Neural Networks, using both the aggregated standard traffic data and the proposed architecture. Different oversampling methodologies are analyzed to balance the training data, finding that the Deep Convolutional Generative Adversarial Networks technique with random undersampling presents better results when generating synthetic instances that permit maximizing the predictive performance. Computational experiments suggest that this model outperforms other traditional prediction methodologies in terms of AUC and sensitivity values over a range of false positives with greater applicability in real life.
The wine industry faces a highly competitive environment, making cost-effective management of the wine supply chain essential. Literature has shown that this objective can be achieved with the ...implementation of horizontal collaboration strategies in logistics. In this strategy, firms located at the same level of the supply chain cooperate to reduce costs, improve quality of service and mitigate environmental externalities. This paper analyses the implementation impacts of a horizontal collaboration policy in the wine supply chain. To do so, we propose a cooperative game with transferable costs, in which the characteristic function is obtained by solving a novel linear programming formulation that models the joint planning of the wine supply chain. To evaluate the benefits of collaboration, we conduct a case study involving three of Chile's largest wineries. The results show that the use of collaborative frameworks leads to significant reductions in the logistics costs of the wine supply chain. Furthermore, we find that the grand coalition reduces the costs by 10.17% compared to the non-collaborative case. This reduction comes mainly from a decrease in the bulk wine inventory cost. We also analyse the impact of coordination costs on the savings and conduct a sensitivity analysis.
Electromobility in public transport has become a promising way to reduce environmental pollution. Several contributions have sought to estimate the energy consumption of buses in public transport. ...However, most of these efforts use measurements collected from controlled or simulated experiments, or that do not characterize the entire bus network. Unlike these studies, this article estimates the energy consumption of all the electric buses that circulate in the city of Santiago, Chile, during the studied period using full disaggregated GPS data and empirical measurements on some sensorized electric buses. The methodology considers a feature selection phase and the development of energy consumption prediction models using physics based and machine learning approaches. The performances of both models are compared with each other, and then, the best one is used to measure the impact of electromobility in the city. This analysis allows decision-makers to target investment by determining the buses with higher energy consumption savings in the face of budget constraints.
•Estimation and impacts of public buses energy consumption.•Comparison of energy consumption prediction of physics and machine learning approaches.•Quantification of energy savings by replacing diesel buses based on SVR consumption estimates.•Sensorized data provides better compared to the use of only the data available for the entire network.
•We present a multi-objective MIP to support wine grape harvest.•We propose a negotiation protocol to ease an agreement between two decision-makers.•Our approach draws from the augmented e-constraint ...method.•We suggest a substitution rate criteria to chose which solution to present.•The protocol might be extended to other contexts with two decision-makers.
In this paper, we present a novel multi-objective mixed-integer linear programming model to support wine grape harvesting. The proposed model considers the opposing nature of operational cost minimization and grape quality maximization, subject to several constraints, such as grape requirements and routing decisions. Based on the operations of a winery we worked with, we develop a negotiation protocol that can lead to an agreed final harvest schedule. The protocol includes an initial Pareto optimal solution obtained through the augmented weighted Tchebycheff method. Then, the solutions are presented to the two decision-makers and, if no agreement is reached, we conduct an iterative process, which includes finding Pareto optimal solutions in a neighborhood using the augmented ∊-constraint method. Finally, we choose, within this set, the solution following a substitution rate criteria. We illustrate our procedure using an educational example.
The COVID-19 pandemic strongly affected the mobility of people. Several studies have quantified these changes, for example, measuring the effectiveness of quarantine measures and calculating the ...decrease in the use of public transport. Regarding the latter, however, a low level of understanding persists as to how the pandemic affected the distribution of trip purposes, hindering the design of policies aimed at increasing the demand for public transport in a post-pandemic era. To address this gap, in this article, we study how the purposes of trips made by public transport evolved during the COVID-19 pandemic in the city of Santiago, Chile. For this, we develop an XGBoost model using the latest available origin-destination survey as input. The calibrated model is applied to the information from smart payment cards during one week in 2018, 2020, and 2021. The results show that during the week of maximum restriction, that is, during 2020, the distribution of trips by purpose varied considerably, with the proportion of trips to work increasing, recreational trips decreasing, and trips for health purposes remaining unchanged. In sociodemographic terms, in the higher-income communes, the decrease in the proportion of trips for work purposes was much greater than that in the communes with lower income. Finally, with the gradual return to in-person activities in 2021, the distribution of trip purposes returned to values similar to those before the pandemic, although with a lower total amount, which suggests that unless relevant measures are taken, the low use of public transportation could be permanent.
Understanding the mobility of surface freight transportation is relevant in urban planning and for developing public policies. Literature shows that most previous efforts on this topic rely on ...surveys and limited data. In contrast to other works, in this paper, we present an innovative methodology for characterizing last-mile freight transportation that uses a novel and copious data source: mobile phone data, which provides a broader scope. Our methodology involves calibrating supervised machine-learning models that allow us to link cell phones with truck drivers. In this endeavor, we construct several input variables that track mobile phone’s daily movement patterns, including traveled distances, interactions with highway networks, and land use variables. We test our approach by conducting a case study in Santiago, Chile, for which we analyze mobility patterns and logistics indicators disaggregated by day, hour, and zoning. For this case, we show that all supervised models performed well regarding AUC, which can be attributed to the high granularity and handling of the data. However, we chose to use NGBoost in all subsequent experiments, as it provided slightly better results on our validation data. Our work has several implications for practice. For instance, our results can support decision-makers and policymakers in identifying critical areas where urban logistics centers and transportation interventions are needed. Finally, several research lines stem from our work, which include assessing the impact of incorporating geospatial information and the measurement of logistics sprawl over time.
•We study VMS effectiveness by proposing a novel vehicle-by-vehicle approach.•We use full-real world data gathered from Automatic Vehicle Identification technology.•87.50% (71.85%) of the messages ...failed to induce speed reduction (lane change)•Heavy vehicle and low-mileage drivers are more likely to follow lane change messages.•Drivers are less likely to obey messages indicating far incidents or road works.
Variable Message Signs (VMS) provide real-time information on traffic conditions, making it possible to guide drivers through electronic signs along the road. Relevant literature has proved VMS to be effective, especially for diverting traffic during incidents in the highway or inducing a speed reduction. Previous efforts, however, usually involve off-highway experiences, including the use of simulators or stated preference surveys, or the measurement of aggregate values of traffic through technologies that are prone to a higher failure rate, such as loop detectors. For bridging this gap, in this research, we propose a novel vehicle-by-vehicle approach (VBV), that differentiate by vehicle type, to assess the impact of VMS on drivers’ road behavior patterns along a section of a Chilean urban highway during risky situations. In addition to the messaging information, we use full traffic data obtained from free-flow gates equipped with automatic vehicle identification (AVI) technology. We conduct statistical analyses to study two potential messaging-induced behavioral changes, namely speed reduction and lane changes. For the speed reduction behavior, in 87.50% of the studied messages, the results indicate that the messages failed to induce the desired change in behavior. This value decreases to 71.85% for lane changes. The results indicate that heavy vehicle drivers and low-mileage drivers are more likely to follow lane change messages.
The study of vehicle stops in last-mile delivery has gained ground in the specialized logistics literature. An efficient last-mile delivery reduces distribution costs and mitigates negative ...externalities such as pollution and congestion. This paper estimates the stops of last-mile trucks that deliver food products in Santiago, Chile. The aim is to study last-mile delivery operations using a non-intrusive, low-cost method. Particularly, we devise a novel methodology that employs multiple data sources to detect the primary stops of cargo vehicles. The proposed methodology involves the following two steps. First, we use GPS data to identify all the candidates for stops, that is, clusters of points close to each other in terms of distance and time, and then, these stops are classified as primary according to the proximity to planned visits or check-out markings. Finally, we conduct a case study involving food distribution, deriving managerial and public policy insights. We find that the most adequate time threshold to detect stops in our context is 4 min, which is considerably lower than previous studies. This last may be explained by the last-mile nature of our study. Our results show that primary (i.e., delivery) stops are concentrated mainly in the center of Santiago, with a duration that decreases as the hours go by. This last means that some of the externalities caused by truck stops (e.g., road capacity reduction) are exacerbated during the morning rush hour. We also find that the average duration of the primary stops is 12.5 min, while the mean distance traveled between two consecutive stops is 4.68 km.
