Image: MKFotônica
Published on 09/08/2025
By Roseli Andrion | Agência FAPESP – Innovation does not always come from a giant leap forward. Sometimes, it emerges from a detour or the simplification of a complex idea. This is how a precise and persistent refinement of a physics technology began to take shape at an innovation hub in Campinas, in the interior of the state of São Paulo, Brazil. It is already emerging as a promising option with high potential for impact.
The optoacoustic sensor can detect gas, oil, or vapor leaks in critical industrial environments, such as places with high radiation, explosion risk, or extreme temperatures. The technology works without touching the fluid, does not require electricity at the measurement point, and can operate safely above 200 °C.
MKFotônica developed the solution with support from the FAPESP Innovative Research in Small Businesses Program (PIPE). It captures mechanical vibrations and ultrasonic noises from faults in industrial equipment with high sensitivity. Instead of a complex integrated chip, its architecture uses discrete components that are already available on the market. “The solution is less elegant but more viable,” summarizes Leandro Matiolli, physicist and co-founder of the startup.
The sensors are based on purely optical principles and are immune to electromagnetic interference, making them ideal for environments where traditional electronic options are ineffective, says Matiolli. “Since they have no electrical parts, they’re safe for critical environments.” The sensors detect frequencies up to 500 kHz, allowing for the early identification of faults, even before they become noticeable to humans or other sensors. Thus, the technology can make monitoring critical industrial processes safer and more efficient.
Optical fiber and simplification
The device consists of an optoacoustic sensor and a reading unit, or optical interrogator, which processes signals from mechanical vibrations on the monitored surface. The interrogator allows specific frequency ranges to be selected, which makes it possible to eliminate background noise and recognize acoustic patterns associated with leaks, wear, or imminent failures.
Thus, the startup offers sensitive, robust technology that can help solve a problem that costs Brazilian companies millions: identifying dangerous leaks before they become catastrophes. The initial idea, however, was quite different.
The original plan was to design a photonic chip, the heart of the system. However, difficulties soon arose, including high cost, complex packaging, and the need for high sales volume to ensure economic viability. “We wanted to develop an optical interrogator for acoustic sensing with a photonic chip. During testing, we realized that, although technically interesting, it wouldn’t be economically viable,” Matiolli explains.
In addition, temperature variations affected the length of the optical fibers, compromising the accuracy of the solution. “We made machines to wind the fibers and control their length with micrometric precision. With thermal variation, however, the system didn’t behave as expected and the interrogator lost functionality.”
Experience led to entrepreneurship
MKFotônica was born out of a change in professional direction. After working in the optical telecommunications industry for a decade – the last four as a development manager – Matiolli decided to become an entrepreneur. He had already participated in creating photonic chips and followed the process of transforming this technology into a product. “I always wanted to start a company. I wanted to transform technical knowledge into practical, accessible solutions,” he says. “When the idea of developing this component arose, I felt that the time was right.”
For this project, Matiolli found the ideal partner in Bernardo de Barros Correia Kyotoku, a fellow physicist and specialist in integrated photonics. “We met at the Telecommunications Research and Development Center [CPQD, an information and communication technology innovation center based in Campinas], where we gained technical experience. We always talked about entrepreneurship and applying our knowledge to market solutions. Ten years later, we founded the company together,” Matiolli recalls.
Initially, they planned to monitor internal electrical discharges in power transformers in the energy industry. However, during the development of the project, the scientists encountered conservatism in the sector. “We talked to engineers, technicians, managers, and everyone thought the idea was excellent, but there were no plans to invest in new equipment. They prioritize data analysis using existing sensors.”
They then researched the petrochemical industry, where leaks in valves, pipes, and ducts pose operational, environmental, and financial risks. “We realized that if we simplified the sensor, we could serve this market. This change was motivated by conversations with Nestor Moura, one of Brazil’s most experienced experts in acoustic emissions. He showed us real problems in the industry and saw great potential in our technology.”
Ironically, simplification made the project more innovative: the quest to detect noises associated with leaks in valves or steam traps, common in high-pressure industrial processes. This is possible because the wide range of detectable frequencies, combined with acoustic pattern identification algorithms, ensures that the application recognizes subtle signs of failure in the early stages – something essential for safety and accident prevention.
In addition to operating in hostile environments, the sensor can be installed up to 30 kilometers (km) from the optical interrogator. The interrogator is fully passive, meaning it requires no local power supply, and has no electronic components at the measurement point. It is compatible with SCADA platforms, which are standard in industrial automation, as well as USB and cloud computing. “The device produces no sparks or heat and is immune to electrical noise. This makes it ideal for explosive environments, substations, high-voltage lines, and other installations.”
Since it can operate safely in classified areas, the technology can be used to measure acoustic emissions, identify partial discharges in transformers, and operate in steel processes and even nuclear reactors because it is immune to radioactive environments. “This reduces risks by allowing constant monitoring.”
Known technology, new application
Although optical sensors have been around since the 1980s, the concept used by MKFotônica has been applied in a creative and functional way to a real contemporary need. “Perhaps this is the right time to apply it, in a niche that requires robustness and reliability. The demand exists, and our optical solution, combined with new machine learning techniques, meets it,” the physicist reflects.
Like many other innovations that only became popular when the market was ready for them, fiber optic leak detection may be finding its place in the Industry 4.0 era. Sensors “hear” the invisible, and algorithms “understand what is said” to transform this information into safety alerts. “With the digitization of industry, integration with real-time monitoring systems makes all the difference.”
In this case, it is worth remembering that the innovation goes beyond the device itself. It is the intelligent application of a known technology to current market needs. “Our sensor follows this logic. It isn’t the most modern, but it’s robust and prepared for extreme environments. Thus, it’s the most useful for what’s needed now.”
Artificial intelligence
One of the main challenges currently facing the researchers at MKFotônica is using the system in noisy industrial environments. While the sensor showed extremely high sensitivity to small leaks in the laboratory, background noise in the field makes it difficult to read the signals clearly. The startup will now incorporate machine learning algorithms into the device to optimize it. “The idea is to train models to identify the sound signatures of leaks amid other sounds. Artificial intelligence can help us extract information accurately.”
It is expected that the algorithms will enable the identification of acoustic patterns of different types of failure. This will reduce false alarms and improve operational efficiency. “Over time, the system learns the ambient sound, and any deviation, even minimal, can be an indication that something isn’t right. This is crucial for those who take care of maintenance,” he ponders. According to Matiolli, continuous analysis of sound patterns will enable predictive decision-making based on reliable information and actionable interpretations of the data obtained. “The ability to monitor in the ultrasound range allows us to identify specific high-frequency signatures. The ability to filter low frequencies ensures the elimination of ambient noise and facilitates the detection of leaks. The combination of filtering and data interpretation capabilities brings an interesting advantage to this technique.”
Ongoing partnerships
The startup already has several partnerships in progress. One of these is with a company in the steel industry and seeks to detect leakage risks and increase operational safety at critical points where there is a risk of explosion. “This company has already tested other sensors, but none of them worked in their environment,” he comments. “It’s a big challenge, but it could become a product directly in their line.” Details are kept confidential to prevent competitors from gaining an advantage.
MKFotônica is committed to national technology and adapting to different industrial sectors. The company is betting on the concrete application of science. “We don’t just sell a sensor; we offer an intelligent way to listen to and interpret what happens at the most critical points in factories,” says Matiolli. “It’s a tool that transforms corrective maintenance into predictive maintenance. A robust, sensitive, and adaptable solution with high-level national technology that solves real problems.”
Source: https://agencia.fapesp.br/55807
