Artificial intelligence to detect of metalworking machines anomalies
July, 22 at 02:45 P.M.
Zyfra, Finnish-Russian industrial digitalization leader, has developed predictive analytics solution powered by machine learning to enhance the capabilities of metalworking machines by detecting of anomalies in technological process and identifying their possible cause.
Zyfra PdA solution analyses the real-time data from CNC machines, alerts users whenever an anomaly emerges, and indicates the possible cause while giving recommendations for further actions. The system classifies a range of anomalies like tool quality, operator error in case of incorrect operating modes selection and machine failure.
“Manufacturers of large-dimensioned products made of expensive materials face a detect defection problem. Defects are caused by many factors such as the quality of the cutting tools, technological faults, equipment wear, etc. There is a need for an intelligent system which will take into account all the necessary information from equipment monitoring systems, evaluate the impact of this information on defects and help operators and technologists to make decisions,” said Alexander Smolensky, Business Development Director of Zyfra.
Metalworking machines failures might lead to defects in manufactured parts or cause equipment to break down. Implementation of the system results in cost reduction of manufactured parts, improvement in product quality and decline in spoiled products amount, as well as reduction of maintenance and repair costs and equipment downtime.
According to Vason Bourne independent research firm, over 80 percent of companies have experienced an unexpected outage within the past three years. Nearly three-fourths of organizations say zero unplanned downtime is now a top priority or the No. 1 priority for their company. The average cost of unplanned equipment downtime is $260,000 per hour, according to research conducted by Aberdeen.
Implementation of Zyfra PdA is envisioned for CNC machines embedded with MDCplus real-time machine monitoring and manufacturing data collection system. The operating principle of the equipment monitoring system is that each machine automatically transfers data about its own performance into a single digital system. Data about the equipment’s condition and workload, and its operating mode, are sent from the machines to computers and other devices equipped with special software, making it possible to promptly eliminate downtime and providing an objective evaluation of the quality of machine operators’ work.
Zyfra has connected 10,000 CNC Machines to its MDCplus real-time machine monitoring and manufacturing data collection system. The projects have been implemented in Bulgaria, China, Finland, France, India, Romania, Turkey and Singapore. India has become the key foreign market for Zyfra with more than 500 CNC machines connected to MDCplus system in 2019. By 2021 the company is looking at more than 2,000 MDCplus installations in India.
Zyfra started its operations in September 2017 in Helsinki, Finland. The company develops industrial digitalization technologies for machinery, metallurgy, mining and oil & gas. The solutions include predictive analytics and data analysis, tech processes optimization, automatic dispatch systems, autonomous dump trucks and teleoperated equipment.
Zyfra’s earnings in 2019 amounted to more than 50 million dollars.