Petrozavodsk Researchers Teach Mining Equipment to “Hear” When It Is About to Fail
At a crushed-stone plant, crushers and vibrating screens process rock under heavy loads. If a component inside a machine breaks – a bearing, shaft or mounting – the entire production line can grind to a halt. That means lost production, financial losses and emergency repairs.
Industrial equipment “speaks” through vibrations. When a machine is running smoothly, its vibration signature remains steady. But when an imbalance develops, bearings wear out or components become misaligned, the vibration pattern changes. Researchers at the Institute of Physics and Technology of Petrozavodsk State University (PetrSU) have developed a compact hardware and software device that detects these changes long before a machine breaks down.
A Compact Device That Can Prevent Costly Repairs
The device attaches to the machine housing. An accelerometer measures the intensity and direction of its vibrations. A microcontroller then performs spectral analysis, breaking down the vibration signal into its component frequencies. Each machine has its own vibration “signature”: a bearing, for example, produces one frequency, while an imbalance produces another. If a particular frequency becomes too strong, it can indicate a fault. When a computer breaks a vibration signal down into frequencies, however, it works with discrete segments of the signal rather than a continuous wave. This can introduce errors. To address the problem, PetrSU researchers combined two interpolation methods – techniques for smoothing the data between measured points – reducing the error in amplitude measurements from 37% to 6.1%. The result is an inexpensive device that delivers highly accurate measurements.
The sensor transmits processed findings over Wi-Fi using Modbus TCP, an industrial communications protocol that serves as a standard language for factory equipment. Instead of sending raw data, it delivers ready-to-use diagnostic results. Maintenance personnel can review an alert, verify the issue and schedule repairs, replacing a component during planned downtime rather than after a breakdown.
The device was tested at stone-processing facilities in the Republic of Karelia, including sites in the city of Suoyarvi and the settlements of Kvartsitny and Pukhta. The sensors recorded operating characteristics of jaw crushers and helped identify problems with how the equipment was being operated.
Unexpected Breakdowns Become Less Likely
According to Dmitry Ivliev, head of the ESTLS Group of Companies, businesses that invest in predictive analytics increase annual profits by 21%, reduce operating costs by 17% and achieve profitability 23% higher than their competitors. Digital tools help companies detect changes in equipment condition in time, maintain traceable records of failure history through a centralized registry with end-to-end analytics, including AI-based analysis, and shorten the time between an alert and a management decision.
In 2023, Petrozavodskmash, part of Rosatom’s Atomenergomash division, deployed an automated vibration-monitoring system for its machine tools, developed by PetrSU. Sensors installed on electric motors, hydraulic power units and machine-tool spindle assemblies monitor vibration, temperature and electrical current characteristics. Neural-network algorithms predict equipment failures. The hardware cost was five times lower than that of comparable systems.
At the time, the Petrozavodskmash system relied on centralized data processing and neural-network algorithms running on an external computing platform. The sensors developed for mining companies perform spectral analysis themselves and transmit processed results, eliminating the need for a separate server.
TIUS has also implemented several wireless vibration-diagnostics projects at TechnoNICOL plants. For example, last year, 28 vibration sensors were installed on seven machines at the company’s facility in Chelyabinsk. The system transmits data using the LoRaWAN protocol, with measurements taken every 15 minutes. The deployment reduced repair costs and equipment downtime by a factor of two to three.
For Operations With Limited Budgets
The PetrSU technology is particularly relevant to jaw crushers, vibrating screens and other equipment used to process mined rock. These machines operate under substantial mechanical loads, and a failure in a single critical component can bring an entire crushed-stone production line to a standstill.
The compact sensor could make advanced diagnostic methods accessible even to companies that cannot afford expensive centralized systems requiring a dedicated server and extensive cabling infrastructure.
