AC arc fault detection enabled through edge AI for arc fault circuit interrupters
Our solution provides a cost-effective and reliable way to detect and interrupt faults, meeting the stringent safety standards of UL1699 and IEC62606
Application overview
Arc Fault Circuit Interrupters (AFCIs) play a vital role in preventing electrical fires by detecting hazardous arcs and interrupting power supply before damage occurs. However, traditional AFCIs often rely on outdated analog circuits and basic signal processing, which can lead to complex signal acquisition requirements and potentially fail to detect arcs before they ignite a fire.
In contrast, our innovative approach leverages a proprietary Convolutional Neural Network (CNN) model trained on extensive load and arcing signature datasets, expertly optimized to run on the TinyEngine? NPU, which is integrated into the?MSPM0G5187 microcontroller family.
This cutting-edge technology enables more accurate and reliable arc detection, setting a new standard for AFCI performance.
Starting evaluation
Data collection
Utilize the TIDA-010971 reference design that provides an unified platform design to enable reliable acquisition of AC arc faults with live AC current data and arc voltage data.
Data quality assessment
High-quality training data is the backbone of any successful AI model. TI’s Goodness of Fit (GoF) evaluation ensures your dataset represents a wide range of real-world motion scenarios from subtle movements to environmental changes. Our Goodness of Fit (GOF) evaluation checks that your dataset captures real-world arc-fault conditions for accurate training. Explore GOF or use our sample dataset in?CCStudio? Edge AI Studio.
Figure 1.1 - Poor data quality
Figure 1.2 -?Well-separated feature clusters for arc vs non-arc
GoF test illustrating the influence of pre-processing feature extraction steps on data separability.
Build and train your model
With your dataset ready, explore, train, and evaluate models using the Edge AI Studio software development tool for edge AI. This intuitive GUI-based tool simplifies the entire process of model creation and deployment.
Get started quickly with the PIR motion detection example project, complete with a preloaded dataset.
Prefer custom data? The platform includes integrated data capture and hosting tools within? Edge AI Studio.
Model Flexibility to support your application needs based on the memory constraints and compute requirements.
Deploying your model
CCStudio? Edge AI Studio?gives a start to finish workflow for deploying trained models to embedded targets. For developers seeking deeper customisation and control, refer to the deployment guides that offer a comprehensive framework for building and integrating EdgeAI functionality into your own embedded applications
- Deploy machine learning models on MSPM0 microcontrollers using EdgeAI Studio GUI Tools
- Deploy machine learning models on MSPM0 microcontrollers using CLI tools
Choosing the right device for you
| Product number | Processing core | NPU available | Clock frequency (MHz) | ? ? ? ?Arc fault model metrics | ||
|---|---|---|---|---|---|---|
| Latency (ms) | Flash (kB) | SRAM (kB) | ||||
| MSPM0G5187 | Arm? Cortex?-M0+ Core | Yes | 80 | 0.39 | 5.55 | 2.4 |
| AM13E230x | Arm? Cortex?-?M33? | Yes | 200 | 0.2 | 5.8 | 2.4 |
All the hardware, software and resources you’ll need to get started
Hardware
LP-MSPM0G5187
MSPM0G5187 LaunchPad? development kit evaluation module,?This is needed for the EdgeAI inferencing, current data is captured using ADC and passed to the AI model after feature extraction.
TIDA 010971
TI reference design to capture AC current data for ARC fault
Software & development tools
CCStudio? Edge AI Studio
A fully integrated no-code solution for training and compiling PIR Motion detection models, to deploy onto TI embedded microcontroller devices.?
CLI tools
Use this end-to-end model development tool that contains dataset handling, model training and compilation?
MSPM0-SDK
The MSPM0 SDK provides the ultimate collection of software, tools and documentation to accelerate the development of applications for the MSPM0 MCU platform?under a single software package.?