Flash News

To Know FFB Maturity by AI Photo: Accuracy to be Perfect

Special doc./Artificial Intelligence (AI) to know fresh fruit bunch maturity
To Know FFB Maturity by AI Photo: Accuracy to be Perfect

InfoSAWIT, JAKARTA – One latest innovation in palm oil industries was the application of artificial intelligence (AI) technology. It could detect fresh fruit bunch (FFB) maturity level and calculate FFB numbers in gathering spot automatically.

As InfoSAWIT quoted from the Linkedin of Prasetyo Mimboro, Plant Investment, Replanting and GIS, PT. Perkebunan Nusantara IV (Persero), every photo or field documentation in geo-tagging which means, it has location information or coordinates, the technology significantly delivered advantages, such as, the increasing harvest more efficiently, FFB maturity could be automatically detected and fast. This would minimize the dependence to man’s effort and accelerate harvest process. The smallholders could also identify FFB to be harvest in much more efficient ways.

Prasetyo also noted, the AI system also classified FFB based on its maturity with high accuracy. This would help the fault to determine harvest period and confirm that only mature FFB would be harvested.

“Production optimization by knowing and understanding when FFB is mature optimally, plantation management should plan the harvest period better. This would help optimize crude palm oil (CPO) production and reduce the loss of resource,” he wrote.

It is believed the application of the technology would minimize the losses because of early detection to rotten FFB or too mature FFB. The management could act faster to avoid the rotten FFB and maximize the harvest.

Accurate area map from geotagging would enable the geographic location sign in the data, including the information about plantation and tree condition. By combining geotagging data with geographic information system, the plantation management would map the plantations more accurate. This information would also help plan area use, supervision, and get decision better.

Real time – geotagging would enable to supervise plantation condition and plants. By using sensor and satellite technology, it would access actual data about the plants growth, moisture, and other environmental factors. These would help optimize the production and minimize the losses.

Field worker verification in geotagging is about to verify the field work, such as, population census, production, topography map, and infrastructure inventory. Geotagging data would visually deliver evidences that can be verified and used to get decision.

The most important thing is that there would be less human erro. Geotagging would minimize human error to track down the location and monitoring. Data that got geographic sign would be more reliable and accurate. “The monitoring process to transport (FFB) would inform the gathering spot location, the numbers and would smooth the transportation process, and minimize the leftover,” he noted

The accuracy of the technology is nearly perfect. As Prasetyo noted, the last accuracy reached 99,34%. “Until now we keep adding our dataset,” he wrote.

The innovation would escalate not only efficiency and accuracy in the harvest process but also deliver strategic advantages to get area and production management in a whole. By implementing AI technology and geotagging, palm oil industries would realize much higher productivity level and better cultivation. (T2)