Edge Impulse Experts / AI-Based Mechanical Anomaly Detector (Audio) Public

Target: Cortex-M4F 80MHz Clone this project

Welcome to Edge Impulse, the largest community of edge AI developers!
This is a public Edge Impulse project, use the navigation bar to see all data and models in this project; or clone to retrain or deploy to any edge device.

AI-Based Mechanical Anomaly Detector (Audio)

Audio

BLE PCB mechanical anomaly industrial Beetle ESP32-C3

About this project

This neural network model detects sound-based mechanical anomalies:

normal
anomaly

After building my neural network model, I deployed my model as a fully optimized and customizable Arduino library and uploaded it to Beetle ESP32-C3. Also, I developed an Android application from scratch to notify the user of the detected anomalies.

Data (28 samples)

View all

ANOMALY_01_11_2024_06_23_13

ID: 805308958

Label: anomaly

Sample length: 8s

NORMAL_01_11_2024_06_13_37

ID: 805308971

Label: normal

Sample length: 3s

ANOMALY_01_11_2024_06_23_06

ID: 805308961

Label: anomaly

Sample length: 6s

NORMAL_01_11_2024_06_13_32

ID: 805308972

Label: normal

Sample length: 4s

ANOMALY_01_11_2024_06_23_00

ID: 805308963

Label: anomaly

Sample length: 5s

ANOMALY_01_11_2024_06_22_28

ID: 805308967

Label: anomaly

Sample length: 4s

NORMAL_01_11_2024_06_13_52

ID: 805308979

Label: normal

Sample length: 3s

ANOMALY_01_11_2024_06_22_40

ID: 805308962

Label: anomaly

Sample length: 4s

Model accuracy (Unoptimized (float32))

Validation set

100.0%

Test set

100.0%

On-device performance (Cortex-M4F 80MHz, Quantized (int8))

Latency

376 ms.

Peak RAM usage

35.2K

Flash usage

32.9K

Run this model

Scan QR code or launch in browser

Launch in browser

Dataset summary

Data collected

2m 8s

Sensor

audio @ 8KHz

Labels

anomaly, normal

Project info

Project ID	336608
Project version	1
License	Apache 2.0
No. of views	6,461
No. of clones	4