This study aims to assess delivery lockers as a last-mile solution, considering the behavior and interaction among e-commerce stakeholders. In Brazil, a large portion of deliveries are attended home. This results in a large number of failed deliveries and redelivery attempts, which can reach up to three delivery attempts. Delivery lockers (DL) represent a solution to reduce the number of re-deliveries and consolidate the delivery of goods. The implementation of DL and the exclusion of the third delivery attempt are evaluated in this study via an agent-based simulation model (ABSM) in the city of Belo Horizonte, Brazil. Four scenarios were simulated by varying delivery locker implementation and the exclusion of the third attempts, which allowed comparable results to be obtained for each agent, in terms of gains and operational and external costs (emission, noise, and congestion). The scenario that most closely resembles the current situation was the one that presented the worst results. The implementation of the lockers brings improvements to this scenario in terms of reducing re-deliveries and distances traveled by trucks. When the possibility of three delivery attempts is excluded, the benefits are even greater. The deployment of lockers also enables carriers to reduce the number of trucks required to make deliveries, thereby increasing their profits.

This work simulated several alternatives for the dynamic allocation of additional human resources in a company that produces a group of specific products. The goal was to increase the average amount of the margin of the total contribution through a hybrid application of a discrete event simulation (DES) and an agent-based modelling simulation (ABMS). Two different decision-making forms were proposed to determine which workstation should receive an additional operator. The first proposal was based on the occupancy level of the operators, while the second one was based on the intermediate queue size. The computational model was operationally validated by comparing the results with actual production data from the company. Twelve scenarios were analysed using a margin of the established contribution. Based on the occupancy rate, the ratio improved on average by 27.68 %, with an additional operator in the workstation. According to the second criterion, this improvement raised to 117.51 %.

This work seeks to study one of the most complex and important issues in production scheduling research: flexible job shop systems. These systems are extremely important for industry, which uses the make-to-order strategy and seeks mix and volume flexibility. The model system will use agents within discrete-event simulation models, generating a Hybrid Simulation model. The agent will sequence the production orders at the beginning of the process and re-sequence them, when necessary, in order to achieve a multi-objective optimization. For this, the agent will bring together two different logics that have opposing goals. This work consists of the comparison of the results of three scheduling methods: firstly, with the sequence of arrival; secondly, with the agent using one sequencing logic; and, finally, using the same logic, but with adjustments in the sequence during the batch production, seeking to improve the negative points generated by the logic. It also stresses that this schedule ensures that the Manager Agent reduces makespan and increases machine utilization while increasing its interference in the model. This is a quantitative study, using the modeling and simulation method and following an empirical model.

The load and weight process in a cement industry is one of logistic step that shows the biggest time of occurrence, increasing the queues. This study aims to do scenarios to solve this queue problem. This way, it pretends to find an better resources distribution.

No abstract