Product Details
Target Applications
The XENO+ TML120 Tiny ML Module is designed for various edge AI applications, including:
- Noise Suppression & Sound Classification for reducing background noise.
- Human Alert Detection for recognizing alarms, glass breaking, and loud noises.
- Machine Vibration & Industrial Monitoring to analyze machine sound patterns.
- Human Voice Command Recognition for voice-controlled applications.
- Health Monitoring for heartbeats, SPO2 analysis, and panic scream detection.
- Automotive Safety for car crash sound detection and SOS alerts.
- Smart Buildings for HVAC and air handling anomaly detection.
- I2C Sensor Data Processing for medical and industrial applications.
Block Diagram and Video
Specifications
| No | Feature | Specification |
|---|---|---|
| 1. | Part Number | XN0-TML120-H - 3.3V Variant XN0-TML120-L - 1.8V Variant |
| 2. | Form Factor | Solderable Module (LGA Pads) |
| 3. | Dimensions | 15mm x 15mm |
| 4. | Operating Temperature | -40°C to 85°C |
| 5. | Power Supply | 3.3V or 1.8V Variants Available |
| 6. | Host Interface | Serial UART |
| 7. | Programming Interface | UART (PC Flashing Tool) |
| 8. | AI Processing | 6.4 GOPS/s |
| 9. | Supported ML Architectures | CNN, RNN, LSTM, GRU, Convolution, Pooling |
| 10. | Neural Parameters | Up to 7M+ parameters (1-bit mode) |
| 11. | Audio Interfaces | PDM, I2S, TDM (Input & Output) |
| 12. | Sensor Interface | I2C for ML Applications |
Electrical & Power Ratings
| No | Parameter | Min | Typ | Max |
|---|---|---|---|---|
| 1. | Supply Voltage (VDD) | 1.8V | 3.3V | 3.6V |
| 2. | Power Consumption | - | - | 20mA |
| 3. | Operating Temperature | -40°C | 25°C | 85°C |
Technical Documentation
| S.no | Title | Document Link |
|---|---|---|
| 1. | Product Brochure | Click Here to View |
| 2. | Use Case | Click Here to View |
Design & Development
Hardware Development
TML120 - XENO+Tiny ML Module EVK Kit
The TML120 - XENO+ Tiny ML Module EVK Kit is designed to accelerate the development of edge ML-enabled IoT devices. It empowers developers, researchers, product designers, and students to rapidly prototype and deploy AI/ML applications using the powerful TML120 module. The EVK Kit simplifies development with a PC Flashing Tool for programming and configuring the ML processor via the Programming UART Port. The Host UART Port streams ML classification results to the host processor, enabling real-time monitoring using tools like Putty or Teraterm.
Model Development
Human Alert Model
One of the key features of the XENO AI/ML device is its pretrained model availability. Here we have a pest sense model which can be a variety of pest sound classification. Here we have a different model which can detect the sound of pests and can be used for research and development purposes
Pest Sense Model
One of the key features of the XENO AI/ML device is its pretrained model availability. Here we have a Human Alert model which consists of 3 different classes for sound classification. Based on the sounds detected it can alert users and necessary action can be taken.
Vibration Sensor Model
Applying Edge ML to vibration data from accelerometer and gyroscope sensors is increasingly valuable for anomaly detection, predictive maintenance, leak detection, and condition monitoring of industrial machines and motors. We provide a reference ML model for classifying anomalies and failure patterns from vibration data, which customers can adopt and fine tune to meet their specific application requirements.
Hardware Development
TML120 - XENO+Tiny ML Module EVK Kit
The TML120 - XENO+ Tiny ML Module EVK Kit is designed to accelerate the development of edge ML-enabled IoT devices. It empowers developers, researchers, product designers, and students to rapidly prototype and deploy AI/ML applications using the powerful TML120 module. The EVK Kit simplifies development with a PC Flashing Tool for programming and configuring the ML processor via the Programming UART Port. The Host UART Port streams ML classification results to the host processor, enabling real-time monitoring using tools like Putty or Teraterm.
Model Development
Human Alert Mode
One of the key features of the XENO AI/ML device is its pretrained model availability. Here we have a pest sense model which can be a variety of pest sound classification. Here we have a different model which can detect the sound of pests and can be used for research and development purposes
Pest Sense Mode
One of the key features of the XENO AI/ML device is its pretrained model availability. Here we have a Human Alert model which consists of 3 different classes for sound classification. Based on the sounds detected it can alert users and necessary action can be taken.
Vibration Sensor Model
Applying Edge ML to vibration data from accelerometer and gyroscope sensors is increasingly valuable for anomaly detection, predictive maintenance, leak detection, and condition monitoring of industrial machines and motors. We provide a reference ML model for classifying anomalies and failure patterns from vibration data, which customers can adopt and fine tune to meet their specific application requirements.