Imprimir Republish


Determinants of survival in industry

Experts assess the risks and opportunities created by potentially disruptive technological developments

Economists from the Federal University of Rio de Janeiro (UFRJ) and the University of Campinas (UNICAMP) are conducting an assessment commissioned by the Brazilian Industry Confederation (CNI) on the impacts from a set of highly disruptive emerging technologies on industrial competitiveness over a horizon of 10 years. The study will provide insights to inform planning in industry and policy-making in government. The project, dubbed Industry 2027, is due to be completed in April, but preliminary findings reveal an outlook that is both troubling and challenging: it is troubling in that Brazil’s standing in the technology race is far from comfortable—out of 759 large and medium-sized firms responding to the survey, only 1.6% are at the technological frontier of advanced manufacturing or Industry 4.0, with integrated, connected, and intelligent manufacturing processes (see chart on page 30); and it is challenging in that 21.8% of those firms said they hope to catch up with the frontier by 2027, which will require research and development (R&D) funding, organizational change, and public policy support.

Forty researchers in different fields are contributing to the program and exploring how and to what extent new technology will determine industrial competitiveness. One of the findings is that none of the 10 industries covered by the survey—including agribusiness, oil & gas, and automotive (see chart on page 30)—will be immune to the disruption brought about by a set of eight key technology clusters: IoT (Internet of Things); communication networks; cloud computing, Big Data and artificial intelligence; connected and intelligent production; bioprocessing and biotechnology; advanced materials; nanotechnology; and energy storage.

The experts running the study believe, however, that there is still time to prepare for the transition. These technologies are likely to be immediately disruptive only in some industries, such as advanced materials in aerospace, and artificial intelligence in the capital goods industry. In most sectors, the transformation will be gradual. “Much of the impact is yet to come. These findings have an upside for firms that are planning ahead, because they show there is still time to develop and fine-tune their strategies,” says David Kupfer, a professor at the Institute of Economics at UFRJ and one of the coordinators of Industry 2027. He notes that while technologies related to artificial intelligence and nanostructured materials will require immediate attention, in most segments technological change over the next 10 years will be incremental. “Early in the survey, we were especially concerned, for example, with constraints on power supply and the high costs of energy storage, but expert analyses have since shown that the transformation will occur incrementally over the next 10 years and there is no immediate risk.” Even in such a competition-critical field for agribusiness as biotechnology, abrupt changes are not likely in the coming years.

In most industries the transformation will be gradual, allowing time to develop and refine strategies

João Carlos Ferraz, also a professor at the Institute of Economics at UFRJ, says one of the goals of the program is to broaden discussion about the future of local industry. “Brazil has a diverse and heterogeneous industry landscape and in general the focus has been only on the most advanced firms. Our survey is concerned with exploring the planning implications for companies at different stages of technology adoption and making appropriate recommendations on government policy,” he says.

The study shows that the eight technology clusters will likely converge to augment each other’s impact. Artificial intelligence algorithms, for example, can be incorporated in cybersecurity solutions in combination with communication network technologies, or can be used in tandem with synthetic biotechnology to analyze genetic data. Innovation in artificial intelligence, IoT, and connected production are expected to cause disruptive impacts by 2027 in 9 of the 10 covered industries. Machines with sensors will communicate with each other. “They will generate large amounts of data about performance and, with artificial intelligence, they will learn how to make their own decisions,” says Antônio Carlos Gravato Bordeaux, a consultant responsible for the IoT component of the study and a director at technology and innovation consulting firm BXTEC. Value chains will be connected, linking companies to their suppliers. “The Internet of Things will improve the way companies manage their employees by allowing them to monitor factory floors posing safety risks and helping to prevent injuries,” says Bordeaux, who served as innovation management director at the Telecommunications Research and Development Center (CPqD) in Campinas from 2001 to 2013. The study estimates that IoT can increase productivity by as much as 50% depending on the industry.

As fast as the new technologies are developing, their costs are declining (see opposite). “This is expanding adoption,” says João Carlos Ferraz. Data from Business Insider show that the average cost of IoT sensors is around US$0.44 per unit, down from US$1.30 in 2004. Cheaper sensors can also aid the transition to a circular economy. “After manufacturing a product, a company can track it throughout its useful life and ensure it is recycled back into a raw material,” says Bordeaux.

Cognitive capabilities
The study predicts scenarios in which the pace of change in industry could accelerate if machines gain advanced cognitive capabilities, but also points to barriers such as difficulties encountered in deploying new technologies seamlessly across all links in the value chain. The build-up of innovation over time could lead to radical change. New, disruptive business models are likely to be created around bundled products and services. “Servitization will tend to become increasingly mainstream,” says Bordeaux. “Instead of buying a refrigerator, consumers will pay for the use of an IoT-connected unit monitored by the manufacturer, and will have it replaced with the next generation if they so choose.”

Another promising front is digital twins, or virtual representations of production lines operating in parallel with the real-world factory floor, allowing engineers to test changes to improve efficiency and safety. “They are already a reality, for example, in the aerospace industry, where products need to be of the highest quality and part wear needs to be accurately predicted.” According to Bordeaux, industry will need investment in technology adoption and more data scientists to work with artificial intelligence. He notes the importance of setting up IoT and advanced manufacturing testbeds to demonstrate potential applications to prospective adopters.

Information streams
Whereas the roadmap for IoT and artificial intelligence in industry is quite a tangible one, another related technology cluster—communication networks—still needs to move away from the commonly held perception of it as just another element of infrastructure. Communication networks are systems of computers, transmission channels, and related interconnected devices through which information is exchanged. “They are needed to tie together the different digital technologies, collecting data from IoT sensors, transmitting information for processing in data analytics and artificial intelligence systems, and conveying information streams in intelligent production,” explains physicist Claudio de Almeida Loural, a former telecommunications researcher at CPqD who advised the program in the field ​​of network technologies.

Although there are currently some mature technologies in this field, such as fiber optics and previous-generation mobile networks, other technologies are still in the selection phase. Examples of these include machine-machine communication standards and protocols, which will compete with each other as they develop to maturity. Another example Loural describes is product connectivity. “The focus is now on energy-efficient, long-range networks. A variety of protocols have been developed or are under development and several of them are likely to exist side-by-side in the coming years.” Most manufacturing companies will likely be cautious when it comes to investment. “Investing in networks is capital-intensive and they run the risk of choosing a protocol that might not gain traction in the end,” he says. Loural sees small and medium-sized businesses in Brazil as especially vulnerable. “Most lack the infrastructure capabilities and knowledge about information and communication technologies to make the necessary leap and leverage emerging network technologies,” he says.

Multiple technologies are not a problem for the energy storage cluster. “There is no single, dominant technology that can be claimed to be better than the rest, as it depends on the application,” says Roberto Torresi from the Institute of Chemistry at the University of São Paulo (USP), who advised the program on energy storage. Among the most mature technologies using electrochemical methods for energy storage are automotive lead-acid batteries, lithium-ion batteries used in electronics and electric cars, and fuel cells, with the latter still developing toward economic feasibility. Lithium-ion batteries are increasingly cheap to produce and have attracted large streams of investment globally (see Pesquisa FAPESP, issue no. 258). The reason, of course, is that batteries will play a crucial role in the development and mainstreaming of electric vehicles as well as energy storage systems to buffer fluctuations in power supply from renewable sources, such as wind, and more conventional renewables like hydro.

Impacts from this cluster will largely be moderate over the next 10 years—although in the automotive industry they are already disruptive. But progress in IoT innovation will be constrained by the need for new technology to provide the power needed to operate sensors and drones. According to Torresi, Brazil is far from being a significant producer of technology in this field, but new opportunities could arise if local electric utilities begin to invest in storage.

The coming years should see increasingly widespread use of high-performance materials such as carbon nanotubes, 3D printing materials, and biopolymers in manufacturing. These developments are being assessed in two different technology clusters within the program: nanotechnology and advanced materials. In nanotechnology, the study has found that while Brazil has strong academic capabilities, R&D efforts in the country have been outstripped by those of competitors. “I was surprised to find that Brazil’s scientific output in nanotechnology, accounting for only about 2% of global output, is less than the country’s 2.7% share of overall R&D output,” says Osvaldo Novais of the USP São Carlos Institute of Physics, who led the nanotechnology component of the program. “We have quite a large community working in this field and have had considerable investment in the last 15 years.” According to Novais, Brazil is lagging in the global effort to use nanotechnology to develop new cancer treatments, microelectronics components, and sensors. “We need aggressive initiatives so we don’t lose ground, especially in the life sciences industry, where prospects are very promising.”

Advanced manufacturing depends on technology convergence and a new level of institutional collaboration, says Kupfer

Innovative companies
The R&D effort needed to address technological challenges in industry will require more intricate institutional collaboration than exists in Brazil today, says David Kupfer of UFRJ. “Advanced manufacturing is dependent not on radical innovation, but on technology convergence. It requires internet integration, investment in Big Data, data analytics, and decision-making capabilities,” he explains. “Technological convergence will require institutional convergence. We’re dealing with a very broad spectrum of business applications, and this has made our current approaches to funding innovation obsolete. An institutional redesign will be needed to support the integrated development of these technologies.” This challenge, he says, is even more complex than developing the technologies proper. “Countries such as China and Germany are reformulating entire government departments to accommodate the challenges facing industry and technology. In Brazil, the innovation framework is fragmented.” This is exacerbated by a lack of investment. “We are starting from the bottom in terms of both funding and institutional integration.” The field survey showed that businesses are not oblivious to the problem. “They’re keenly concerned about future-proofing their technology, but an organized effort to drive the development they need is yet to be seen,” he says. “In any case, it will take more than an effort by companies alone, but a systemic one spanning regulatory reform, new services and digital solutions, and integration with the supply chain”.

Researcher Antonio José Felix de Carvalho, from the USP São Carlos School of Engineering, who headed the Advanced Materials chapter of the program, says Brazil’s biggest challenge is expanding the number of innovative companies with the capacity to transform these materials into value-added products. “We are competitive in basic and low-tech materials, but have virtually no firms developing materials for other high-tech industries. Look at the challenges involved in producing electric vehicles. They require special steels, high-performance batteries, motors with high-performance magnets, and lightweight polymers to reduce body weight. But there are no companies in Brazil developing these materials,” he says.

Carvalho believes Brazil needs to target industries where it can be competitive, such as oil, wood pulp, and renewable energy, and invest in industrial capabilities involving new materials. Biorefineries, for example, can produce advanced inputs linked to the production of bioenergy. There is a real risk, according to Carvalho, of entire industries losing competitiveness. “The danger is in going the way of the textile industry, which foundered when competitors began developing products incorporating new materials.”

Advances in biotechnology and bioprocessing promise impact in segments such as agribusiness, life sciences, and the environment. “These developments will disrupt these industries, multiplying returns for companies that are keeping ahead of technology,” says biologist Carlos Alberto Moreira Filho, a researcher at USP’s School of Medicine and head of the program’s biotechnology chapter. “DNA editing will enable scientists to develop agricultural species that are resistant to pests and environmental stresses,” he says referring to CRISPR/Cas9, a technique that is much less costly than developing transgenic species. “This is a real game changer. And it relies not only on molecular biology, but also on microchemistry and Big Data. Brazil already has a competitive advantage from its climate and the experience of a country with more than 40 million hectares of transgenic plants.” In health care and medicine, disruption will occur in the production of vaccines, discovery of the molecular mechanisms of diseases, and personalized medicine, says Moreira. “Scientific output from our universities is strong but is difficult to monetize because we lack large, locally-based companies to bring them to market. In the pharmaceutical industry, a merger or consolidation of multiple players would be needed.”

Paulo Mól, the Industry 2027 program manager at CNI, says the initiative is helping industry to think about the future. “We’re still discussing past agendas because Brazil is still struggling with major, long-standing issues such as its social security deficit, infrastructure problems, and barriers in the business environment. It’s time we looked at the technologies that will dominate the landscape over the next 10 years,” he says.