Among the many topics studied in the extensive field of Computer Science and Informatics (CCI), artificial intelligence (AI) is one of the most important due to its transversal nature and its foreseeable impact on the future of society. In modern organizations there are currently multiple examples where this specialty of engineering is applied. Today, the CCI is involved, together with other disciplines, in the handling of large volumes of data, the creation of mathematical and computational models, algorithms and intelligent systems, as well as the development of automatic learning techniques, such as the so-called “ deep learning”, based on artificial neural networks. The computer systems developed allow support for decision-making, extracting knowledge, optimizing production processes.
Informatics and Computing Group CCTVal
Viscovery Project (R+D)
The project consists of developing a platform for processing and extracting information in opinion forums that allows trends to be analyzed and novelties to be detected in the opinions of brands, products and organizations. It uses natural language processing techniques on top of big data processing platforms and combines them with topic models to capture dynamics in text. This project is creditor of a FONDEF VIU-16P0085, under the name: Interactive visual browser for huge volumes of text. Student: Pablo Ortega (Executor). Principal Investigator: Marcelo Mendoza.
High Performance Computing and Grid Computing
Much of the research in High Energy Particle Physics begins with the data generated by the LHC, whose volume is enormous: about 200 petabytes on disk and about 250 petabytes on tape. Currently the ATLAS detector generates around 1 petabyte per second, a volume of data that is drastically reduced to around 1 gigabyte per second by means of an electronic device called a “trigger” that allows selecting some specific types of particle collisions. The distribution, storage and processing of this data requires a specialized collaborative computing network called WLCG (“World Wide LHC Computing Grid”), with more than 150 sites around the world. Since 2013, CCTVal has participated in this network with a "Tier 2" category Data Center, one of the three that exist in Latin America. This Data Center is managed by the CCTVal Computing and Informatics group. This powerful infrastructure offers researchers from USM and other universities the possibility of processing high-performance computing (HPC) jobs, including parallel computing. The Data Center is part of the “National Laboratory for High Performance Computing”, NLHPC. Through this collaboration, USM researchers can also make use of the NLHPC's supercomputing facilities.
Tsunami Warning System
This has been one of the projects with the greatest impact at the national level carried out by researchers from the Computing and Informatics group of CCTVal. As is known, Chile is a seismic country and is constantly related to tsunamis. This reality has prompted the search for solutions that allow modeling these natural phenomena and warning the population, some time in advance, of the imminence of an event of this nature.
CCTVal researcher Patricio Catalán, Ph.D. in civil and academic engineering from USM, together with a group of researchers from other institutions such as SHOA and PUC, addressed this problem based on an exhaustive study of the state of the art in this area, based on the experience of countries such as Japan, that share with Chile similar telluric and maritime phenomena.
In a first stage, a system was consolidated with more than 7,000 simulated tsunami scenarios and coverage of the entire national territory. With the progress of the studies, it was possible to develop an early warning system that allows the decision to be made regarding warning the population of the proximity of the arrival of waves to the coastal region produced by the effect of a tsunami. The wave front of these natural phenomena can travel at more than 600 kilometers per hour in the open ocean and, if the rupture zone occurs close to the coast, the first impact can arrive in a few minutes, which requires computational capacity. and extremely fast reaction.
After eight years of research, it was possible to develop the so-called Integrated Tsunami Warning and Prediction System (SIPAT), which today is one of the most advanced systems in its category in the world and is already part of the Hydrographic System and Oceanographic of the Chilean Navy (SHOA). Among its qualities, the speed and precision of its calculations and scenario projections stand out, which gives the authorities time to order a rapid and focused evacuation of the population in the event of a tsunami on a technical basis. In recent years, the Computing and Informatics group, led by Dr. Luis Salinas, has maintained collaborations with more than thirty universities around the world, and has developed projects for institutions and companies both nationally and internationally. At the same time, the group has hosted important conferences in the specialty and popular talks and symposiums.
BioMed – HPC and Medical Imaging
Project developed in conjunction with the Faculty of Medicine of the University of Chile, focused on the local and remote processing of images of a biological or medical nature, for the purpose of pure research or medical diagnosis. The project consults the accumulation of experience in storage, transmission and handling of large volumes of data, and the use of HPC and computer networks to carry out remote processing, on the web, of images and constitutes an important step in the development of Computational Medicine. The project promotes interdisciplinary work in areas such as computing and informatics, biomedical sciences, biophysics, neurosciences, etc.
Oceana is an international organization focused on the conservation of maritime resources. CCTVal, together with Oceana, presented a project to the Government of Chile that seeks to protect the maritime environment of the Salas y Gómez Islands, through a Remote Surveillance System whose purpose is to monitor and control, in an effective and timely manner, the areas Marine Protected Areas (AMP's) that have been defined as National Marine Parks.
Mining Cluster Project
The project deals with a knowledge management system related to mining supplier companies through a georeferenced technological platform with the aim of measuring, developing, issuing reports and analyzing the territorial competitiveness of companies in the mining sector.
INIA Project (Agricultural Research Institute)
Through the development of genomic and genetic engineering tools, the aim is to have an efficient and fast program that can enhance the genetic improvement of the vine and thus obtain new varieties of table grapes.
The CCTVal, through the CTI-HPC, periodically teaches training courses in various areas of Computing and Informatics, including Satellite Image Processing courses, GPU-CUDA Language courses, etc. In the development of these activities, the CCTVal has the support of the Office of Naval Research Global, (ONRG), the firm nVidia and others. The CTI-HPC was declared Cuda Teaching Center 2012 by nVidia.