AI carrot measurement, counting and packing automation
Agriculture Sector
Challenge
The process of manually measuring, counting, and packing carrots is a labor-intensive and time-consuming task for personnel. It involves the workers manually handling the carrots, measuring and counting them, and then packing them into the appropriate packaging based on weight and quantity requirements. This process is often repetitive and monotonous, which can lead to a decrease in productivity and quality control. Additionally, human error can occur during this process, resulting in inaccurate measurements or incorrect packaging, which can impact the overall quality of the product.
Solution
Our solution addresses this challenge by training an AI model using a 2D Nano C2020 camera to detect and estimate the size and weight of carrots. The carrots are tracked on a conveyor belt using conveyor tracking and encoder. Instead of manual measurement and counting, personnel only need to place the carrots in designated pockets, and it will measure and count them. Once the desired package weight is reached, it signals the packaging machine to initiate the packaging process, streamlining the process and reducing the need for manual labor.
Results
- Eliminated manual measurement and counting* — personnel interaction reduced to placing carrots in pockets, with the system handling all weighing and pack decisions
- Packaging accuracy improved* — target weight reached consistently, reducing giveaway and underweight rejects
- Higher throughput* achieved through automated signaling between the vision system and the packaging machine
What does the system measure?
A 2D camera and an AI model detect each carrot on the conveyor and estimate its size and weight while conveyor tracking with an encoder follows it along the belt. Personnel only place carrots in designated pockets; the system measures and counts them and signals the packaging machine when the target pack weight is reached.
What hardware was used?
A single 2D Genie Nano C2020 camera, a conveyor encoder for tracking, and Byte Motion's AI vision software connected to the customer's packaging machine. No robot was needed: the automation replaced manual weighing and counting and triggers packaging automatically.