Skip to content
/ SPARK Public

Smart Parking AI-Driven RZBoard Kit

License

Notifications You must be signed in to change notification settings

Avnet/SPARK

Repository files navigation

SPARK - Smart Parking AI-Driven RZBoard Kit

Welcome to SPARK, the edge-AI solution for smart EV charging and parking lot management. Bootstrapping your Edge AI with RZBoard and the RZV AI SDK.

spark_iotc_landing_page

Overview

SPARK is an effective EV charging prototype that integrates a real-time capable convolutional neural network (CNN) for occupancy detection with the RZBoard V2L. The result: a smart parking lot occupancy/EV charging experience for low cost. This system is designed to automate parking lot analysis, streaming edge compute data to your desired target: an information hub like IoTConnect, or an HDMI output. The result: improved user-experience for EV charging customers and parking lot managers.

Looking for the dataset used to train? Check it out on kaggle.

Features

  • Real-Time Occupancy Detection: Utilizing an efficient CNN, SPARK detects vehicle occupancy in real-time, ensuring accurate information on parking availability.
  • User-Friendly Interface: Easy-to-use interface demo interface: output data to HDMI for simple viewing. Or, integrate with IoTConnect for bootstrapped dashboards.
  • Sustainable Solution: Promotes efficient energy use and supports the growing need for EV infrastructure.
  • Scalable Architecture: Designed to scale: one image could host up to 200 parking spots. Try adding more parking slots and/or images until your FPS requirement isn't met. Then, scale horizontally with more RZBoards connected to your IoTConnect style HUB!
  • IoTConnect Enablement: Optionally, configure the demo with IoTConnect by configuring a secrets.json file in /opt/spark/iot. See secrets-template.json for more info.

Installation

Standalone Installer

Looking to demo the software? Simply download and install from the Release instructions.

Development Installer

Want to build and modify the demo yourself? You'll have to run the build script and/or deploy script from within the RZV AI SDK

./build.sh && ./deploy.sh [board_ip]

Operation

SPARK operation is as simple as clicking a button and dragging some boxes.

operation_demo.mp4

License

This project is licensed under the MIT License

Acknowledgments

  • This project wouldn't be possible without OpenCV and RZV AI SDK!

© 2024 Avnet. All Rights Reserved.