Intelligent vibration sensor with in-sensor AI
17/08/2026
STMicroelectronics NV has introduced an intelligent vibration sensor designed for industrial condition monitoring applications that require high accuracy, reliability and energy efficiency. Built using ST’s microelectromechanical systems (MEMS) technology, the IIS3DWB10IS vibration sensor featuring intelligent sensor processing unit (ISPU) 2.0 enables advanced digital signal processing and artificial intelligence (AI) inference closer to the sensing element. The result is a compact, rugged device that measures vibrations and shocks of up to 200 g at frequencies of 10 kHz and above.
Combining digital precision and ease-of-use with a wide temperature range of up to 125°C that can withstand harsh environments, the vibration sensor is engineered to help customers improve equipment uptime, reduce unplanned downtime and support predictive maintenance strategies across industrial environments. Vibration analysis is claimed to be the dominant segment of condition monitoring, as many industries use rotating and oscillating machinery for cutting, shaping, moving, cooling and other processes. The ability to prevent equipment stoppages through the early detection of issues, such as predicting bearing failures in advance, helps companies across a range of sectors, including automotive and other manufacturing activities, to optimise production flow.
The IIS3DWB10IS vibration sensor is claimed to be the first digital sensor with both a wide bandwidth and embedded processing capability to deliver performance that meets the needs of high-end industrial condition monitoring applications, offering a compelling alternative to piezoelectric sensors. Accurate measurement of vibrations above 10 kHz, with a large dynamic range of up to 200 g, combines with a noise floor as low asa35 μg/Hz, which is comparable to the noise performance of piezoelectric sensors.
Moreover, the IIS3DWB10IS delivers high accuracy and sensitivity, adding digital-sensing advantages, including simplified electrical and mechanical design, smaller size, lower power consumption and greater flexibility in the computational partitioning. The ISPU 2.0 introduces new hardware accelerators to perform real-time signal processing and edge AI. These hardware accelerators make frequently used functions faster and more power efficient.
The core is C-programmable and contains on-chip program and data random-access memory (RAM). The supporting ecosystem provides software libraries that facilitate the execution of typical vibration monitoring algorithms in the ISPU, including fast Fourier transform (FFT), filtering, envelope, velocity severity and anomaly detection. With 40 million instructions per second (MIPS) and 40 mega floating-point operations per second (MFLOPS) digital signal processing, the ISPU 2.0 delivers up to four times the processing performance of the previous generation.
In addition, the ISPU 2.0 sensor interface supports six times faster data transfer with the MEMS circuitry. The IIS3DWB10IS also contains a 2048 × 80-bit first-in, first-out (FIFO) register and an accurate temperature sensor. The sensor’s rugged MEMS-based design supports operation up to 125°C. The IIS3DWB10IS will be supported into the future according to ST’s ten-year industrial longevity programme.
The IIS3DWB10IS is packaged as a 4.5 mm × 4.5 mm × 1.5 mm 16-lead land grid array (LGA) package with wettable flanks that facilitates automatic optical inspection in high-quality surface-mounted assembly processes.
