Plantation Application Software
Grading Sensor
Grading Sensor
Grading Sensor

Fresh Fruit Bunches grading is normally performed manually by competent and experienced graders who collect samples and evaluate the FFB based on visual observation.

FFB grading is considered the front-line assessment in any palm oil mill to measure raw material ripeness and freshness. The quality of the raw material will determine the yield of oil produced at the end of the process, often referred to as OER (Oil Extraction Rate) and Free Fatty Acid (FFA).

However, the sample size is only about 25-30% of the whole loads which means only 50 to 100 bunches are assessed per load. In other words, in terms of quantity, there are more ungraded FFB than there are graded ones for each load.

In addition, most of the samples are picked from the top part of each load. If the quality of the FFB at the bottom portion of the load is significantly different from the top, it means that the grading quality of the load may be inaccurate and can lead to misrepresentation towards the calculated OER of the day.

By utilizing a multi-spectral camera for image capture and a suitable machine learning algorithm, palm oil mills can teach the machine to identify the quality of each FFB which can then be sorted according to their quality. This would be beneficial during the sterilization process where the specific process can be chosen for the sterilization of FFB.

Palm oil mills are using mainly triple-peak sterilization. However, if the FFB can be segregated according to their quality, the sterilization process can be further optimized to operate at an energy-efficient setting, hence reducing the need for triple-peak sterilization operation.

Advance analysis such as this has the potential for better accuracy and can assess 100% of the FFB inside a moving scraper conveyor. By implementing this, manpower can be reduced, and fewer disputes regarding quality and quantity will take place between palm oil estates and mills.

Keywords: FFB Grading Sensors

FFB grading is considered the front-line assessment in any palm oil mill to measure raw material ripeness and freshness. The quality of the raw material will determine the yield of oil produced at the end of the process, often referred to as OER (Oil Extraction Rate) and Free Fatty Acid (FFA).

However, the sample size is only about 25-30% of the whole loads which means only 50 to 100 bunches are assessed per load. In other words, in terms of quantity, there are more ungraded FFB than there are graded ones for each load.

In addition, most of the samples are picked from the top part of each load. If the quality of the FFB at the bottom portion of the load is significantly different from the top, it means that the grading quality of the load may be inaccurate and can lead to misrepresentation towards the calculated OER of the day.

By utilizing a multi-spectral camera for image capture and a suitable machine learning algorithm, palm oil mills can teach the machine to identify the quality of each FFB which can then be sorted according to their quality. This would be beneficial during the sterilization process where the specific process can be chosen for the sterilization of FFB.

Palm oil mills are using mainly triple-peak sterilization. However, if the FFB can be segregated according to their quality, the sterilization process can be further optimized to operate at an energy-efficient setting, hence reducing the need for triple-peak sterilization operation.

Advance analysis such as this has the potential for better accuracy and can assess 100% of the FFB inside a moving scraper conveyor. By implementing this, manpower can be reduced, and fewer disputes regarding quality and quantity will take place between palm oil estates and mills.

Keywords: FFB Grading Sensors

Accelerating Your Processes with Business Process Automation

www.plantation.aiwww.plantation.ai