Plantation, Crop and Wildlife Monitoring Vision
Vision models that grade oil palm fruit, spot crop disease and count or identify plants and animals in the field.
Tier Grades
Tier A: Frontier AI labs
not feasible
Frontier labs do not work on oil palm grading; Google publishes general wildlife and crop models.
Tier B: Indonesian AI labs
feasible
Training took about 3.1 hours on 6,592 images, well within an Indonesian lab budget.
Tier C: Institution units such as ITB
feasible
Indonesian and Malaysian universities already publish these models.
Tier D: Organisations such as SEMAI (PoT, PoC, PoV)
feasible
FFB grading and Ganoderma detection are direct plantation proofs of value; no plantation deployment is stated in the source.
Where this comes from
Tiers: a) frontier AI labs, b) Indonesian AI labs, c) institution units such as ITB, d) organisations such as SEMAI delivering proof of technology, concept and value. Each rationale cites the anchors below.
Anchors
- ComputeA YOLOv8 oil palm FFB grader trained in about 3.1 hours (Gunadarma University).
- Data6,592 oil palm FFB images from Roboflow across ripeness categories; precision 94.5%, recall 94.7%, mAP 98%.
- Team size (proxy)Proxy only (author count of an exemplar paper, not a headcount): 4 authors on the Gunadarma FFB paper (Jurnal Ilmiah Teknologi dan Rekayasa 30(3), 2025).
- Time horizonMonths.
- Shelf-life signalsA plantation reports a production deployment of FFB grading, Drone and satellite models replace ground photos
- Pipeline roleTier C builds datasets and models with plantations; Tier B scales them; Tier D runs grading and disease pilots at mills and estates.
Where this comes from
Team size is a proxy (author counts of exemplar papers), not a headcount.
Who and What
Exemplars
YOLOv8 FFB ripeness grader (Gunadarma)
YOLOv11 FFB ripeness grader (IJCCS 2025)
TF-ResNet50 Ganoderma detector
SpeciesNetObserved actors
Linked models
YOLOv8 FFB ripeness grader (Gunadarma)visionexemplarYOLOv11 FFB ripeness grader (IJCCS 2025)visionexemplarTF-ResNet50 Ganoderma detectorvisionexemplarSpeciesNetvisionexemplar
MegaDescriptorvisionproduced by
NTS-Netvisionproduced by
YOLOv8visionused byResNet50 (ImageNet)visionused byCropNetvisionproduced byCOTS detectorvisionproduced byPlant disease classifier (38 classes)visionproduced byOil palm detector (Universitas Gunadarma, Roboflow)visionproduced byOil-Palm-Health (gunadarma, Roboflow)visionproduced byGanoderma Detection (Oil Palm), BINUS and PTPN IVvisionproduced byOil Palm Ganoderma Detection (binus, Roboflow)visionproduced byPalm oil FFB ripeness (zah, Roboflow)visionproduced byPalm Oil Ripeness (braincore, Roboflow)visionproduced bySawit ripeness (testing workspace, Roboflow)visionproduced byLinks
R&D TopicPlantation, Crop and Wildlife Monitoring Vision
ModelsYOLOv8 FFB ripeness grader (Gunadarma)Gunadarma University, visionTF-ResNet50 Ganoderma detectorMultimedia University, visionYOLOv11 FFB ripeness grader (IJCCS 2025)UPN Veteran Jawa Timur, visionSpeciesNetGoogle, vision+14 more
YOLOv8 FFB ripeness grader (Gunadarma)Gunadarma University, visionTF-ResNet50 Ganoderma detectorMultimedia University, visionYOLOv11 FFB ripeness grader (IJCCS 2025)UPN Veteran Jawa Timur, visionSpeciesNetGoogle, vision+14 more
Models 18
TF-ResNet50 Ganoderma detectorMultimedia University, FGV R&D Sdn Bhd, visionexemplarYOLOv11 FFB ripeness grader (IJCCS 2025)UPN Veteran Jawa Timur, visionexemplarYOLOv8Ultralytics, visionused byYOLOv8 FFB ripeness grader (Gunadarma)Gunadarma University, visionexemplarMegaDescriptorCzech Technical University in Prague, visionproduced byOil palm detector (Universitas Gunadarma, Roboflow)Gunadarma University, visionproduced byResNet50 (ImageNet)visionused byCOTS detectorGoogle, visionproduced byCropNetGoogle, visionproduced byGanoderma Detection (Oil Palm), BINUS and PTPN IVvisionproduced byNTS-NetPeking University, visionproduced byOil Palm Ganoderma Detection (binus, Roboflow)visionproduced by+6
Oil-Palm-Health (gunadarma, Roboflow)visionproduced byPalm oil FFB ripeness (zah, Roboflow)visionproduced byPalm Oil Ripeness (braincore, Roboflow)visionproduced bySawit ripeness (testing workspace, Roboflow)visionproduced bySpeciesNetGoogle, visionexemplarPlant disease classifier (38 classes)visionproduced by
SpeciesNetGoogle, visionexemplarPlant disease classifier (38 classes)visionproduced byPapers
- An open access repository of images on plant health to enable the development of mobile disease diagnosticsDavid. P. Hughes et al., arXiv preprint, 2015 founding
- BioCLIP: A Vision Foundation Model for the Tree of LifeSamuel Stevens et al., CVPR 2024 (per arXiv comment), 2023 state of the art
- Efficient Pipeline for Camera Trap Image ReviewSara Beery et al., KDD19 (per arXiv comment), 2019 state of the art
- The iNaturalist Species Classification and Detection DatasetGrant Van Horn et al., CVPR 2018 (per arXiv comment), 2017 state of the art
- Deep learning in agriculture: A surveyAndreas Kamilaris et al., Computers and Electronics in Agriculture International Journal, 2018, 2018 survey
Where this comes from
The papers to read first: where the direction began, its current best, and a survey when one exists.
Sources
- Optimalisasi Deteksi Tingkat Kematangan Tanda Buah Segar Kelapa Sawit Menggunakan YOLOV8 Dengan Platform Web | Jurnal Ilmiah Teknologi dan R ejournal.gunadarma.ac.id https://ejournal.gunadarma.ac.id/index.php/tekno/article/view/67
To cite this page: Project AURION, AI Use-Cases Research & Innovation, "Plantation, Crop and Wildlife Monitoring Vision", https://aurion.semai.id/topics/plantation-crop-and-wildlife-vision/, records of October 2026.