Abstract: In 2000, Chile introduced profound health reforms to achieve a more equitable and fairer system (GES plan). The reforms established a maximum waiting time between diagnosis and treatment for a set of diseases, described as an opportunity guarantee within the reform. If the maximum waiting time is exceeded, the patient is referred to another (private) facility and receives a voucher to cover the additional expenses. This voucher is paid by the health provider that had to do the procedure, which generally is a public hospital. In general, this reform has improved the service for patients with GES pathologies at the expense of patients with non-GES pathologies. These new conditions create a complicated planning scenario for hospitals, in which the hospital's OR Manager must balance the fulfillment of these opportunity guarantees and the timely service of patients not covered by the guarantee. With the collaboration of the Instituto de Neurocirugia, in Santiago, Chile, we developed a mathematical model based on stochastic dynamic programming to schedule surgeries in order to minimize the cost of referrals to the private sector. Given the large size of the state space, we developed an heuristic to compute good solutions in reasonable time and analyzed its performance. Our experimental results, with both simulated and real data, show that our algorithm performs close to optimum and improves upon the current practice. When we compared the results of our heuristic against those obtained by the hospital's OR manager in a simulation setting with real data, we reduced the overtime from occurring 21% of the time to zero, and the non-GES average waiting list's length from 71 to 58 patients, without worsening the average throughput.

Abstract: Strategic, tactical, and operational harvesting plans for the forestry and logging industry have been widely studied for more than 60 years. Many different settings and specific constraints due to legal, environmental, and operational requirements have been modeled, improving the performance of the harvesting process significantly. During the summer of 2017, Chile suffered from the most massive wildfires in its history, affecting almost half a million hectares, of which nearly half were forests owned by medium and small forestry companies. Some of the stands were burned by intense crown fires, which always spread fast, that burned the foliage and outer layer of the bark but left standing dead trees that could be salvage harvested before insect and decay processes rendered them unusable for commercial purposes. Unlike the typical operational programming models studied in the past, in this setting, companies can make insurance claims on part or all of the burnt forest, whereas the rest of the forest needs to be harvested before it loses its value. This problem is known as the salvage logging problem. The issue also has an important social component when considering medium and small forestry and logging companies: most of their personnel come from local communities, which have already been affected by the fires. Harvesting part of the remaining forest can allow them to keep their jobs longer and, hopefully, leave the company in a better financial situation if the harvesting areas are correctly selected. In this work, we present a novel mixed-integer optimization-based approach to support salvage logging decisions, which helps in the configuration of an operational-level harvesting and workforce assignment plan. Our model takes into account the payment from an insurance claim as well as future income from harvesting the remaining trees. The model also computes an optimal assignment of personnel to the different activities required. The objective is to improve the cash position of the company by the end of the harvest and ensure that the company is paying all its liabilities and maintaining personnel. We show how our model performs compared to the current decisions made by medium and small-sized forestry companies, and we study the specific case of a small forestry company located in Cauquenes, Chile, which used our model to decide its course of action. (C) 2019 Elsevier Ltd. All rights reserved.

Abstract: We are interested in the scheduling problem where there are several different resources that determine the speed at which a job runs and we pay depending on the amount of each resource that we use. This work is an extension of the resource dependent job processing time problem and the energy aware scheduling problems. We develop a new constant factor approximation algorithm for resource cost aware scheduling problems: the objective is to minimize the sum of the total cost of resources and the total weighted completion time in the one machine non-preemptive setting, allowing for arbitrary precedence constraints and release dates. Our algorithm handles general job-dependent resource cost functions. We also analyze the practical performance of our algorithms, showing that it is significantly superior to the theoretical bounds and in fact it is very close to optimal. The analysis is done using simulations and real instances, which are left publicly available for future benchmarks. We also present additional heuristic improvements and we study their performance in other settings. (C) 2018 Elsevier B.V. All rights reserved.

Abstract: The prognostics and health management disciplines provide an efficient solution to improve a system's durability, taking advantage of its lifespan in functionality before a failure appears. Prognostics are performed to estimate the system or subsystem's remaining useful life (RUL). This estimation can be used as a supply in decision-making within maintenance plans and procedures. This work focuses on prognostics by developing a recurrent neural network and a forecasting method called Prophet to measure the performance quality in RUL estimation. We apply this approach to degradation signals, which do not need to be monotonical. Finally, we test our system using data from new generation telescopes in real-world applications.

Abstract: The ever increasing challenges posed by the science projects in astronomy have skyrocketed the complexity of the new generation telescopes. Due to the climate and sky requirements, these high precision instruments are generally located in remote areas, suffering from the harsh environments around it. These modern telescopes not only produce massive amounts of scientific data, but they also generate an enormous amount of operational information. The Atacama Large Millimeter/submillimeter Array (ALMA) is one of these unique instruments, generating more than 50 Gb of operational data every day while functioning in conditions of extreme dryness and altitude. To maintain the array working under extreme conditions, the engineering teams must check over 130,000 monitoring points, combing through the massive datasets produced every day. To make this possible, predictive tools are needed to identify, hopefully beforehand, the occurrence of failures in all the different subsystems.
This work presents a novel fault detection scheme for one of these subsystems, the Intermediate Frequency Processors (IFP). This subsystem is critical to process the information gathered by each antenna and communicate it, reliably, to the correlator for processing. Our approach is based on echo state networks, a configuration of artificial neural networks, used to learn and predict the signal patterns. These patterns are later compared to the actual signal, to identify failure modes. Additional preprocessing techniques were also added since the signal-to-noise ratio of the data used was very low.
The proposed scheme was tested in over seven years of data from 132 IFPs at ALMA, showing an accuracy of over 70%. Furthermore, the detection was done several months earlier, on average, when compared to what human operators did. These results help the maintenance procedures, increasing reliability while reducing humans' exposure to the harsh environment where the antennas are. Although applied to a specific fault, this technique is broad enough to be applied to other types of faults and settings.