Vibration Classification and Anomaly Detection with BrainChip’s Akida

Many predictive maintenance applications can use neural networks to classify vibration data from accelerometers. In this Expert Project we walk you through data collect, model training, and deployment to BrainChip's Akida™ Development Kit Raspberry Pi containing an Akida AKD1000 for neural network acceleration.

Some key highlights of this project include:

• Collecting low latency accelerometer data with a Raspberry Pi with low latency and uploading to Edge Impulse.

• Training both an Akida classification project and a separate anomaly detection project using the same dataset.

• Utilizing the output from those two project combined with Python code to enable Akida for neural network hardware acceleration and the Raspberry Pi for anomaly detection on the same accelerometer data.

center: 0.121976525
edge: 1.9035794e-11
off: 0.790965
on: 0.08705848
/home/ubuntu/brainchip-accelerometer
Loaded runner for "Brainchip / bc-pred-main-anom"
classification:
{'anomaly': -0.4078322649002075}
timing:
{'anomaly': 0, 'classification': 0, 'dsp': 0, 'json': 0, 'stdin': 28}

Find the full documentation here.