You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: content/hardware/06.nicla/boards/nicla-voice/tutorials/glass-break-detector/content.md
+13-3Lines changed: 13 additions & 3 deletions
Original file line number
Diff line number
Diff line change
@@ -23,6 +23,8 @@ hardware:
23
23
24
24
The Nicla Voice and Edge Impulse offer powerful tools for monitoring environments like spaces with glass structures. This system allows timely detection and responses for environments like buildings, vehicles, or specialized facilities. You can create scalable AI solutions for security needs by leveraging these tools.
25
25
26
+

27
+
26
28
This application note will show you how to build and deploy a machine learning model on Nicla Voice with Edge Impulse that detects glass-breaking sounds. The content will cover how audio data is processed and used for training for efficient sound classification.
27
29
28
30
## Goals
@@ -60,7 +62,11 @@ At the end of the application note, you will have all resources and details for
60
62
61
63
## Edge Impulse and Nicla Voice Versatility
62
64
63
-
Edge Impulse and Nicla Voice provide a highly flexible and adaptable solution for various monitoring applications, including glass structure monitoring systems. It can be integrated and trained to operate in:
65
+
Edge Impulse and Nicla Voice provide a highly flexible and adaptable solution for various monitoring applications, including glass structure monitoring systems.
-**Buildings:** Identify glass-breaking incidents in residential, commercial, or industrial properties. Alerts can trigger lockdown protocols, notify security teams, and log incidents in real time.
66
72
@@ -322,7 +328,7 @@ The Edge Impulse CLI provides real-time monitoring and classification results. T
322
328
323
329
Below is an example of a real-time glass break classification result:
You can also use the Arduino IDE's Serial Monitor to observe similar results found within the Edge Impulse CLI.
328
334
@@ -338,6 +344,8 @@ This enhanced functionality will allow you to design flexible safety and monitor
338
344
339
345
The system begins with the Nicla Voice, which runs a trained machine learning model to detect glass-breaking sounds. For this, we will use the model as discussed in the [Machine Learning Model Development](#machine-learning-model-development) section. Upon recognizing such an event, the Nicla Voice sends a notification via Bluetooth® Low Energy (BLE) to the Portenta H7.
340
346
347
+

348
+
341
349
The Portenta H7 is the host device, bridging the detection hardware with the Arduino Cloud. It receives the BLE notification, processes the alert, and updates the cloud dashboard in real-time. Additionally, the Portenta H7 triggers specific actions based on the alert level:
342
350
343
351
-**Security Alert Call:** It is activated upon recognizing the first glass-breaking event. This is to provide an immediate warning that a critical event may have occurred, enabling early intervention.
@@ -759,9 +767,11 @@ In this example, you can view real-time event logs, track BLE connection status,
759
767
-**Connection Status:** The BLE connection status between the Nicla Voice and Portenta H7 is shown, to provide present system operation.
760
768
-**Security Alert and Lockdown Mechanism Indicators:** You can monitor the activation status of the *Security Alert* and *Lockdown* mechanism directly from the dashboard.
761
769
770
+

771
+
762
772
The cloud integration enables you to receive updates and take action from anywhere, making the system more accessible and responsive. The following clip shows a brief insight of the inference process and communication of Nicla Voice and the Portenta H7:
0 commit comments