The Akida AKD1000 Edge AI processor provides OEMs and car makers with a cost effective and robust ability to perform real-time in-vehicle preventative care by running noise and vibration analysis.


Overview of Edge AI implementation for Vibrational Analysis 

Any piece of machinery produces vibrations or sound that represents a healthy state, which changes over time as the equipment ages or is damaged. While the human ear may not be able to distinguish minor changes in sound that indicate an impending problem, the Akida AKD1000 Edge AI processor can provide early diagnostics through real-time analysis of sensor data on problems such as:

Reducing human-related errors, labor costs, and preventing machinery deterioration in real-time provides invaluable progress towards complete safety and reducing maintenance costs.

  • Mechanical-related looseness
  • Component Misalignment
  • Electric motor failure
  • Gearbox faultiness
  • Lubrication issues

The result is a reduction of labor costs, downtime and machinery deterioration, increasing safety and improving customer satisfaction. The Akida AKD1000 Edge AI processor can also monitor the passenger compartment, checking driver alertness with the aim to reduce human error. Privacy is assured without the need to upload images to the internet.

Edge AI Implementation Challenges

The Akida AKD1000 Edge AI processor provides immense value for real-time analysis. By processing the data within the vehicle, it eliminates privacy issues, lack of service whenever the internet drops out, and the latency and bandwidth issues that plague Cloud-Based AI. The Akida AKD1000 Edge AI processor provides high speed neuromorphic processing of sensor data at a low cost, high speed and at very low power consumption.

Akida Neuromorphic IP for Vibrational Analysis

The Future of Predictive Maintenance

The Akida Development Environment (ADE) provides a high-level neural network API to facilitate the development and emulation of Akida neural network models. The ADE is written in C ++ , and largely inspired by the Keras API. The CWRU (Case Western Reserve University) data set was used to build a reference network for vibration analysis. Data from each class originally comprising of more than 480,000 continuous values was collected and divided into segments of 1024 values.

These segments were resized to a 2-D matrix of 32 by 32. The segment-wise division is illustrated in Figure 1. 1000 segments are chosen at random as test data and rest were used for training.

Classification experiments were conducted on the raw time-series data and continuous wavelet transform (CWT) of the time-varying data. Continuous Wavelet Transform is a method to decompose a real signal into a set of elementary waveforms that provide a way to analyze the signal by examining different components related to its wavelets. The Akida model obtained an overall accuracy of 98% for the classification of raw vibration data without any pre-processing and with only 15 epochs for training. In order to investigate the effects of pre-processing instead of using raw data, CWT features were extracted. Without changing any additional parameters, the classification accuracy of the Akida model improved by 1% thus resulting in an overall accuracy of 99%. This is an excellent result that provides a point of reference for application of the Akida Neuromorphic processor in bearing fault detection.

As the world shifts away from human-driven error and AI implementations are continuously providing value in all applications, the Akida Neuromorphic IP bridges the gap between AI-enabled devices and implementational feasibility.  

For organizations, this translates as a new standard of product development and maintenance practices as it allows for: 

  • Real time equipment and machinery monitoring at peak performance
  • Highly configurable to suit a vast range of products across multiple industries
  • Incremental on-chip learning
  • Ability to run large networks at high speed and minimal power consumption
  • Robustness
  • Self -contained Solutions
  • Secure end-to-end classification
Full Product Overview