Artificial Intelligence and Digital Transformation in Engineering
Introduction
The early 21st century features the beginning of digital data storage dominating over analog technologies. Over the past decade, artificial intelligence (AI) modeling has been used in complex pattern recognition problems such as computer vision and natural language processing. We process these models with high-performance computing architectures using graphics processing units (GPUs), which have high processing capacity and low energy consumption compared to conventional processors (CPUs). The recent results of large language models, such as GPT-3 and its conversational interface ChatGPT, have brought about great expectations (and controversy) regarding the potential (and risks) of artificial intelligence in all areas.
In addition to how it calls for new technologies, digital transformation involves a new business perspective and new forms of asset management. As such, it is mostly an organizational and cultural transformation and not simply a technological development. In this context, the training of specialized human resources in digital technologies is vital for the development of our country. The education process in digital technologies is inherently multidisciplinary as it requires integrating different areas of knowledge and identifying specific problems in this industry.
We at COPPE have been pioneers in developing computational solutions for complex engineering problems since the 1970s. We carry out our work in AI and Digital Transformation in Engineering through the interdisciplinary field of Computer Science and Engineering (CSE). CSE allows for studying complex systems and natural phenomena that would be very costly, or even impossible through direct experimentation. It uses computational resources to solve questions that neither theory nor experimentation can solve by themselves. In CSE we also develop specific algorithms and software systems for engineers and scientists that integrate various subjects and multiple size scales. At the heart of the digital revolution, CSE is driven by recent advancements in high-performance computing and AI and has established itself as a way of generating new knowledge and producing innovation in all areas of engineering, science, technology, and our society.
CSE and COPPE
CSE sits at the intersection of mathematics and statistics, computer science, and the fundamental fields of engineering and science. This combination gives rise to a new field, whose characteristics differ from its original constituents. CSE requires knowledge of domains, mathematical modeling, numerical analysis, algorithm development, software implementation, program execution, and results analysis, validation, and visualization.
Our greatest attribute is how we combine our skills in advanced computing with engineering science to solve complex problems. Here at Coppe, CSE has an increasingly important role on a national scale for the following application areas:
Petroleum and Energy
It is imperative in the oil and gas area that we minimize exploration risk. Therefore, the industry progressively demands high-fidelity subsurface simulations. Other energy forms such as wind energy and biomass combustion, for instance, increasingly require high-fidelity simulations, whether to understand where to position wind turbines or even to design more efficient and less polluting burners. Our institution operates in this area through our digital hub.
IoT and Big Data
The evolution of sensor technology and cheaper storage hardware have led to an increasingly greater supply of data in practically all activities and sectors. IoT and Big Data technologies enable the capture, storage, and data-intensive processing that benefit from mathematical models for decision support and predictions.
Robotics and Automation
Computer simulation is a standard practice in the design of advanced engineering systems such as vehicles, machines, or robots. The engineering design process with high-performance computers has a drastically reduced need to build and test prototypes, as well as greater safety and improved ergonomics.
Computer vision
Recent computer vision and pattern recognition algorithms allow us to analyze images with the precision level of a human being. This type of model has an increasing demand for industrial and medical use.
Smart Cities
Accelerated urbanization brings a series of challenges for the sustainable development of cities. The solution involves the massive use of data collected from different sources and sophisticated mathematical models to analyze this data.
Atmosphere Physics and Climate Change
Climate models will require unprecedented resolution and accuracy to understand how global climate change is related to regional events, where the impact on people and the environment is greater.
Medicine and Biology
CSE technologies are quickly becoming indispensable for medical and biological sciences. Simulation plays an important role in the conceptual development of medical devices. Molecular biology.
Chemistry
Computational chemistry (CC) is widely used in academic and industrial research. Calculated molecular structures, for example, are often more reliable than those experimentally determined. Simulations for protein folding prediction have been used in drug development.
Materials
Our challenge in materials research is to create new materials and improve the existing ones through manufacturing and processing so that they have the desired performance and environmental feedback.