Scroll Top


Ultra-Low Power Neuromorphic Front-End Chip for Vibration Analysis

VibroSense, a Tiny AI chip based on the Neuromorphic Analog Signal Processor, performs pre-processing of vibration data at the sensor and addresses the major challenge of Industrial IoT: amount of data to be transmitted, stored and processed. 

VibroSense Product Family

Tire Monitoring

Ultra-low-power neuromorphic chip for smart tires. Recognizes and transmits tire vibration patterns for the road and tire monitoring 

Machine Health

Ultra-low-power neuromorphic chip for condition monitoring of machines. Enables wireless smart sensors with ML applied locally

Tire monitoring

Tire Monitoring System with VibroSense chip enables road and tire condition monitoring in real time with high accuracy

A car equipped with VibroSense is much safer. Road condition monitoring is performed locally on-tire in real time and the alerts reach Advanced Driver Safety System (ADAS) in microseconds, assisting drivers automatically and enhancing  stability and control, particularly in challenging weather conditions.

Although advanced ADAS systems that can monitor various road conditions such as wet or icy surfaces are not widespread at present, the installation of VibroSense alongside a standard TPMS is sufficient to improve safety for the general public.

Vibration sensors typically operate within a frequency bandwidth of several kHz. However, these high-frequency signals generate vast amounts of data that need to be sampled and transmitted for additional processing and to generate alerts for ADAS and other systems.

VibroSense transforms high-volume vibration data into compact patterns that uniquely match specific road conditions (dry, damp, wet, snow, and ice), which are over 4000 times smaller in size, significantly reducing data volume. Despite the reduced size, the information contained in these patterns is still sufficient for reliable road condition detection.

POLYN’s research has showed that 2-axis accelerometer is sufficient for 92% accuracy of a road surface class detection by the NASP AI model.

Feature Description
Road condition monitoring in real-time
Dry/wet/ice/snow surface impacts the tire friction and grip
Road pavement type detection
For example, asphalt/concrete/gravel/soil
Tire tread wear
Evaluate the tire tread condition and detect wear patterns that may affect performance and safety
Loose wheel nut detection
One loose lug nut adds extra pressure on the others and the uneven pressure distribution will make the wheel shake while driving
Wheel imbalance detection
Any signs of imbalance including partial that may only manifest within a narrow speed range

Machine Health monitoring

Sensor Raw Data Pre-Processing with Neural Networks Saves A Fortune!

Machine condition monitoring can be used to achieve the best possible upkeep of industrial equipment. It involves gathering and analyzing data from sensors such as vibration and acoustic. The machines being monitored are equipped with vibration sensors that collect data and transmit it to a local server or cloud for storage and processing.  The vast amount of data generated by vibration sensors during operations makes condition monitoring costly due to the expenses associated with data transmission, storage, and processing.
VibroSense chip solves the problem by reducing the transmitted data volumes and making condition monitoring more accessible and cost-effective on a global scale.

Challenges Answered by VibroSense

1Raw data pre-processing on-sensor reduces data volumes by 4000 – 1000 times

2Transmitting only small data patterns supports narrow-bandwidth long-distance communications

3Processing significantly less data reduces OPEX and TCO of Predictive Maintenance solutions

Vibration-based condition monitoring is a fundamental Predictive Maintenance technique that is used to detect machine failures. By analyzing vibrations, it is possible to identify a range of machinery problems such as shaft unbalance and misalignment, bearing failures, gear wear, cracks, looseness, and more. 

Vibrations can be used not only for monitoring relatively simple mechanisms, but also complex ones such as pumps, engines, or wind turbines. 

However, vibration signals can be intricate, especially in complex machinery operating at varying speeds and loads. The presence of background and measurement noise further complicates signal analysis. 

Leveraging neural networks for signal processing presents a very appealing approach. Neural networks can extract useful information even from very noisy signal, due to the non-linear way they process data. Some deep neural network architectures, as the utilized in VibroSense, prove to be exceptionally well-suited for addressing vibration monitoring challenges. 


Accurate Data processing with Neuromorphic chip

Supports energy-harvesting based designs

Enables LPWA communications

Reduces cloud connectivity costs

Sensor node and infrastructure optimization

Ensures short time-to-market