Mathematics can help companies innovating

By Elton Alisson in Rio de Janeiro  |  Agência FAPESP – Brazilian scientific production in the field of mathematics won worldwide recognition in 2014 when Artur Ávila was awarded the Fields Medal, the highest distinction in the discipline. In 2018, Brazil was promoted by the International Mathematics Union (IMU) to the elite “Group 5” tier of the 11 most developed nations in mathematics research.

Brazil’s Institute of Pure and Applied Mathematics (IMPA), for example, has surpassed the top US universities, such as Harvard, Stanford and Berkeley, in terms of the scientific impact of the research conducted there. More people earn master’s degrees and PhDs in mathematics from IMPA than from any other Brazilian institution, according to its director, Marcelo Viana.

The mathematical knowledge created in Brazil’s various research institutions must now be leveraged by industry to fuel new businesses, foster product, process and service innovations, and drive the creation of qualified jobs, in the view of the participants in the First Workshop on Mathematics and Industry, held on February 13-14, 2020, at IMPA in Rio de Janeiro.

The aim of the workshop was to put industrial firms in touch with the opportunities and solutions offered by academic research. It was organized by IMPA in partnership with the Center for Mathematical Sciences Applied to Industry (CeMEAI), one of the Research, Innovation and Dissemination Centers (RIDCs) funded by FAPESP. The participants included representatives of Petrobras, Equinor, McKinsey, Stone, Hotel Urbano, Via Varejo, Bayer, Shell and Repsol, as well as other companies.

Viana said Brazil’s excellent research institutions mean that it has significant potential to take advantage of the economic benefits to be gained from mathematics.

“In addition to outstanding institutions, Brazil has sufficient experience in mathematics to realize this potential,” he stressed. “And two of these institutions, IMPA and CeMEAI, are represented at this event.”

One of IMPA’s graduates is Artur Ávila, who earned a PhD from the institute and is now a tenured professor there.

“We prefer our graduates to stay in Brazil, but we’re an institution that exports PhD holders to several countries,” Viana said.

The contributions to industry made by mathematicians who have graduated from Brazilian institutions include predicting the behavior of buildings and other structures and calculating flows of crude oil from drilling rigs by means of mathematical modeling, which involves the description of systems or physical phenomena using mathematical concepts and language.

More recently, they have helped predict processes that have no physical description by combining statistics and high-performance computing.

“The ongoing artificial intelligence wave was made possible by the interaction of computing, mathematics and statistics,” said José Alberto Cuminato, head of CeMEAI, adding that all three disciplines must be involved if mathematicians are to successfully collaborate with industry.

“This is why CeMEAI’s name refers to mathematical sciences and not just to mathematics,” he said.

Mathematics and business

Launched in 2011 at the University of São Paulo’s São Carlos Mathematics and Computer Sciences Institute (ICMC-USP), CeMEAI was a trailblazer in stimulating interaction between mathematical sciences and the industrial sector.

Among its inspirations were the UK’s University Consortium for Industrial Numerical Analysis (founded in the 1980s at the University of Oxford, where Cuminato earned his PhD and now called the Knowledge Transfer Network, or KTN), and the Fraunhofer Institute for Industrial Mathematics in Germany.

“We were the first to speak of mathematical sciences as a business in Brazil, at a time when any connection with the corporate world was frowned upon,” Cuminato said. “Over the years, we’ve learned that it’s a very important niche for mathematicians to develop.”

In the past nine years, CeMEAI’s researchers have conducted projects with small, medium-sized and large companies in various segments, from financial services and healthcare to textiles, poultry, aerospace, and oil and gas. At present, their main collaboration is with Petrobras.

“We aim to depend as little as possible on public funding. We’ve conducted two major projects with Petrobras, which has invested the same amount as FAPESP in CeMEAI,” Cuminato said.

One of the projects is designed to improve the reliability of downhole safety valves (DHSVs), which are subsurface components of oilwells that act as a failsafe. They are critical components because they prevent blowouts, which are uncontrolled flows of oil and other pollutants into the environment when the pool of underground oil erupts.

DHSVs are controlled from the surface by operational personnel on the drilling rig, and failures are undesirably frequent. “The purpose of the project is to investigate the causes of these failures and to help suppliers enhance the product’s reliability,” said Francisco Louzada Neto, CeMEAI’s head of technology transfer.

The other project conducted in collaboration with Petrobras focuses on operational planning, test simulation and reliability assurance for Annelida, a centipede-like robot being designed to clear blockages in oil pipelines.

The idea is that the robot will use sensors to detect hydrate and paraffin plugs and remove them by controlled local heating. “In short, our purpose is to provide mathematical and statistical support to the teams who are developing the robot. This mainly involves measuring and guaranteeing its reliability,” Louzada explained.

Joining forces

Petrobras’s partnership with CeMEAI and with other universities and mathematics research institutions in Brazil has enabled the oil company to cut costs and increase its acquisition of reliable data by means of modeling, according to Danilo Colombo, an engineer at CENPES, Petrobras’s research and development center (CENPES).

“Interaction with academia gives us access to groups with multidisciplinary approaches and a focus on data analysis so that the best results can be extracted from simulation testing of the equipment used in drilling and well completion [making wells ready for production]. This equipment is very costly and must not fail or present errors,” Colombo said.

Petrobras and other oil companies in Brazil invest in collaborative research projects with universities in compliance with a law known as the “R&D clause”, which requires 0.5% to 1.0% of gross sales to be invested in research and development.

The regulator for the industry is the National Oil, Gas and Biofuel Agency (ANP), which oversees investment in R&D.

“These investments are made not by government or public entities but directly by oil companies. They’ve increased more than twentyfold compared with 1997, the year in which the clause was enacted,” said Alfredo Renault, ANP’s head of R&D supervision.

“The scope for more partnerships between universities or research institutions and the oil industry is enormous, especially in research efforts to cut costs and enhance operating safety. Project selection depends mainly on applicability.”

Wellspring of talent

One of the main sources of talent in this field is the Brazilian Mathematical Olympiad, organized each year by IMPA with the support of the Brazilian Mathematical Society (SBM). Almost 20 million students and a record number of cities took part in the 2019 edition. Only 16 cities were unable to participate because they have no public or private schools that teach the second stage of primary education (sixth to ninth grades) or secondary education.

“The Brazilian Mathematical Olympiad is an inexhaustible wellspring of talent,” Viana said. “There are countless examples of successful careers built by medalists in academia and elsewhere. Only a minority pursue an academic career, which is as it should be.”