AutoChef
AutoChef is a revolutionary smart device designed to seamlessly manage your food inventory. It sends timely notifications about expiration dates and suggests personalized recipes to help you minimize waste, optimize meal planning, and maximize savings. Utilizing the Qualcomm® RB3 Gen 2 development kit, an advanced edge device, AutoChef employs AI software trained to recognize and log food items. This ensures you are alerted when and how to prepare the foods you have stored. Additionally, the device integrates with your VR headset to provide real-time information, including camera streams and sensor data, enhancing your culinary experience.
Objective
Food waste is a significant but frequently underestimated issue, with one-third of all food going uneaten in the United States . This not only has a tremendous environmental impact but also causes enormous financial loss—the average American household loses nearly $1,500 in food waste each year. Given the scope of this problem, we were inspired to create a solution that not only saves people money but also contributes to a more sustainable food supply.
Our goal is to drastically reduce food waste at the household level by providing a smart, user-friendly solution. We envision a device that monitors food in refrigerators and pantries, tracks expiration dates, and sends timely notifications to users. To make food management even easier, the device will offer guided recipes based on available ingredients, helping users make the most of what they already have. By implementing this system, we aim to empower individuals to make informed food choices, reduce waste, save money, and contribute to a more sustainable planet.
Tools
| Part Name | Description / Link to purchase |
| Qualcomm RB3 Gen 2 Dev Kit-Vision Kit, QCS6490 | Qualcomm® development kit with Qualcomm® QIRP SDK for on board processing and camera Streaming |
| Meta Quest 3 Device | FVR Headset from Meta |
Source Code
| Description | Link |
| Model Dataset with an optimization to use specific YOLO versions | https://universe.roboflow.com/freshness-detection-vwn1a/freshness-detection-rhrze/model/3 |
| Qualcomm model used as a base layer for model distillation in creation of our model | https://aihub.qualcomm.com/models/yolov8_det_quantized?chipsets=QCS6490+%28Proxy%29 |
| Qualcomm QIRP Camera Demo Example | https://docs.qualcomm.com/bundle/publicresource/topics/80-65220-2/qrb-ros-camera_5_2_5.html?product=1601111740013886 |
| Unity ROS Plugin | https://github.com/Unity-Technologies/ROS-TCP-Endpoint |
| Qualcomm QIRP SDK | https://www.qualcomm.com/developer/software/qualcomm-intelligent-robotics-product-sdk |
| ShapesXR Space | https://shapes.app/space/view/1bebc65a-956d-4545-9019-38146edb958b/23s2d51a |
Additional Resources
| Resource Title | Link |
| Presentation | https://docs.google.com/presentation/d/1i7o-s_hT7I1B-ugdGHgYZIf2y6VPSsXrpXbTsnat9IQ/edit?usp=sharing |
| Devpost | https://devpost.com/software/autochef-3r1ezc |
| Attribute sources | https://www.usda.gov/about-food/food-safety/food-loss-and-waste http://usda.gov/foodlossandwaste/consumers |
Build / Assembly Instructions
1. Block Diagram
2. Connect to the Qualcomm RB3 Gen2 dev kit using adb from the host system using usb
adb devices
adb shell
3. Use nmcli to connect to the wifi
nmcli dev wifi connect <WiFi-SSID> password <WiFi-password>4. Source the qirp-setup.sh
source /usr/share/qirp-setup.sh5. Launch the camera node
ros2 launch qrb_ros_camera qrb_ros_camera_launch.py6. Set the same domain name for the host and the dev kit
export ros_domain_id = 0Note: On the host system you should have access to the topics from the Qualcomm RB3 Gen2 dev kit if not check the firewall settings of the host system.
7. Setup the ROS TCP Connector on the host system and Unity
8. Provide the address and necessary settings and enter the ip address of the host system
9. Create custom ROS publishers and subscribers to communicate with the Qualcomm RB3 Gen2 dev kit
10. Create the build for android and deploy on the VR headset
Usage Instructions
1. Autochef service technicians can provide seamless installation.
2. This project is designed to run autonomously so no work will be done for the end user
3. UX App displaying food data in the fridge
