Online monitoring system prevents unplanned equipment downtime

A combination of wireless sensors and artificial intelligence, developed by a FAPESP-supported company, can help companies of different sizes avoid losses due to production interruptions; the startup took part in a business mission during FAPESP Week Spain.

By Roseli Andrion  |  Agência FAPESP – A combination of wireless sensors and artificial intelligence can help companies of different sizes avoid losses caused by unplanned downtime of rotating equipment. This is the proposal for a system developed by Techplus Automação, a startup based in Campinas, São Paulo state, Brazil. Techplus received support from FAPESP’s Innovative Research in Small Businesses Program (PIPE) to develop a solution that continuously monitors machines to diagnose and predict failures before they occur.

The difference is in the way the monitoring is done. Traditionally, companies use two types of maintenance approaches: corrective maintenance, where repair is done only after failure, and traditional offline predictive maintenance, which relies on periodic visits by technicians to take measurements and assess the condition of equipment.

Traditional offline predictive maintenance compares current and previous measurements to determine if conditions are deteriorating. There is always a risk of problems occurring between these assessments. In addition, many machines are located in hard-to-reach places, which poses risks for technicians and affects worker safety.

After observing this scenario, the company developed a tool to monitor the machines online. “Continuous online monitoring eliminates the need for someone to take measurements in person. This is a double win for the company: it reduces occupational health risks and increases efficiency because the machine’s status is measured continuously,” explains electrical engineer Samarone Ruas, managing director of Techplus Automação.

The resulting data is sent to the cloud and analyzed by artificial intelligence algorithms that alert the company to possible anomalies in the equipment. “Measurement takes place, for example, every minute or every five minutes, depending on the need,” explains Ruas. “This makes monitoring much more efficient and safer.”

The project was entirely designed by Techplus Automação. “We created everything, from the sensor to the algorithm that analyzes the data, including the process of sending the information to the cloud. That’s why the solution is complete: from the electronics embedded in the sensor to the cloud processing with artificial intelligence.”

According to Ruas, the demand for online predictive maintenance solutions has grown significantly over the past two years, as digital transformation and artificial intelligence have advanced in industries. “Every company has rotating machinery: it could be for cooling, pumping, ventilation or compression, for example. Our motto is ‘keep the machines running’ because they literally move the world,” he says.

From small to large companies

One of the system’s key features is its scalability: the solution can be installed in small businesses as well as large industries, such as sugar and alcohol mills or pulp and paper factories, which may need thousands of sensors in a single unit. “You can start with as few as ten sensors, for example. Sometimes a piece of equipment uses more than one sensor, so it may be possible to adopt the system with five pieces of equipment.”

Ruas points out that a tool that predicts the occurrence of faults can mean a 20% to 30% gain in productivity and savings. “The return on investment is between three and six months,” he says. “When we can prove this with numbers, the cost barrier falls easily: companies start to see the solution as an opportunity to reduce costs and increase operational efficiency.”

The expert points out that similar companies are already emerging in the United States, Israel and Germany, but the market is still relatively new. “The adoption of technology usually starts with large companies, the early adopters of digital transformation, and then spreads to medium and small companies – not least because the price goes down when the costs are shared by a larger number of customers.”

For Ruas, adoption of the technology is a gradual process, with companies typically starting with pilot projects before expanding implementation. “We predict exponential growth in the next five years, with the potential to continue for the next ten years,” he concludes.

International markets

Techplus Automação has been automating industries for 20 years. In the last five years, the team realized that digital transformation was already happening in companies. “We saw that there was a large potential market, especially with the advancement of the use of artificial intelligence in this segment.”

The expert points out that predictive maintenance was still in its infancy in industries at the time. “What’s more, the cost was higher than traditional predictive maintenance. Companies thought it wasn’t the right time to invest in this technology because it was more expensive and the method they were using was working,” he recalls. “A couple of years ago, that changed a lot: with digitalization and artificial intelligence, it became easier to prove that it’s an opportunity to reduce costs and increase operational efficiency.”

The company currently serves all of Brazil and already has plans for international expansion in the next two to three years. “The Brazilian industrial market has unfortunately shrunk in recent years and now represents less than 5% of the global market,” he points out. “Because the product is in the cloud and easy to install, the barriers to reaching other markets are low. We see huge potential in Europe, the United States and China.”

Ruas says that the practice of traditional predictive maintenance is similar all over the world. “Companies are still starting to implement online predictive analytics solutions. Today, there are specialized companies that go to industries to take these measurements offline,” he says. “The trend, however, is to replace them with online sensors. We’ve been approached by these traditional companies looking for partnerships to present our solution to their customers.”

The company was one of four deep techs selected to participate in a business mission during FAPESP Week Spain, which took place between November 27 and 28 at the Complutense University of Madrid (UCM) Medical School campus in the Spanish capital.

During the mission, Ruas visited Spanish companies interested in partnerships to market the online monitoring system. “I made contact with two companies that intend to enter into commercial partnerships and that could supply some special types of sensors that could improve the solution we’ve developed,” said Ruas.

“My main objectives during this mission were to establish strategic partnerships with universities, local companies and research centers, as well as to gain a better understanding of the challenges and particularities of the European market” (read more at agencia.fapesp.br/53452).