Firdausa Sonna Anggara Resta

Firdausa Sonna Anggara Resta

Auto-generated profile, based on publicly available information

Firdausa Sonna Anggara Resta

Department of Electrical Engineering, Institut Teknologi Sepuluh Nopember, Surabaya, Indonesia

Research Focus & Metrics

Academic Career Length

1Year (2026-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
18 March 2026 in IEEE Access
Verified
Journal Article
Open access
FRFirdausa Sonna Anggara Resta
Firdausa Sonna Anggara Resta
RSRachmad Setiawan
Rachmad Setiawan
MRMuhammad Rivai
Muhammad Rivai
RARezki El Arif
Rezki El Arif
ANAzis Natawijaya
Azis Natawijaya
AHAfan Ghafar Al Hadad
Afan Ghafar Al Hadad
This work presents a multimodal radar-vision framework for non-destructive ripeness classification of oil palm fresh fruit bunches (FFBs) under field conditions. A continuous wave 2.4 GHz radar and a digital scale capture backscattered power and bunch mass as internal cues related to dielectric properties and growth, while RGB images provide external color and texture information. Radar-weight features (power, weight, and power-to-weight ratio) are modeled using an extreme gradient boosting (XGBoost) classifier, and a YOLOv8-Nano detector is trained using an expert-annotated RGB dataset to localize and classify bunches into four maturity stages (Unripe, Underripe, Ripe, Overripe). Decision-level late fusion combines the class-probability outputs of both branches through mean, weighted, and stacking schemes. Experiments on radar-weight data (255 samples per class) acquired at a commercial loading ramp show that the unimodal radar-weight and vision models reach 89.9% and 94.3% accuracy, respectively, with YOLOv8-Nano achieving mAP@0.5 of 0.99. The best fusion configuration, based on stacking, attains 96.9% accuracy and 97.3% F1-macro, demonstrating that radar-weight information provides complementary evidence that stabilizes vision-based grading and supports more consistent harvest decisions.

23 March 2026 in IEEE Access
Verified
Journal Article
Open access
AHAfan Ghafar Al Hadad
Afan Ghafar Al Hadad
RSRachmad Setiawan
Rachmad Setiawan
TMTotok Mujiono
Totok Mujiono
AAAbdul Latief Abadi
Abdul Latief Abadi
YSYohana Avelia Sandy
Yohana Avelia Sandy
RARezki El Arif
Rezki El Arif
Basal Stem Rot (BSR) caused by Ganoderma boninense is a primary threat to oil palm productivity and is difficult to detect early because visual symptoms appear late. This study presents a detection and severity-classification framework for healthy, mild, and severe based on dielectric characteristics measured ex vivo with a Vector Network Analyzer (VNA) and a Dielectric Assessment Kit (DAK) open-ended coaxial probe over 4 MHz–3 GHz with open-short-load (OSL) calibration. Four dielectric parameters, namely dielectric constant (εᵣ′), dielectric loss factor (εᵣ″), loss tangent (tan δ), and conductivity (σ), were summarized into 20 features per sample using simple statistics, then learned by three Extreme Gradient Boosting (XGBoost) classifiers specialized per plant part: leaf, frond, and root. An all parts combined model was also evaluated as an additional baseline. A balanced dataset of 1,080 samples with 360 samples per plant part was collected at Pusat Penelitian Kelapa Sawit (PPKS) Marihat. Models were trained with an 80% training and 20% test split, using stratified five-fold cross-validation on the training set for hyperparameter tuning. On the held-out test data, macro-F1 reached 0.903 for leaf, 0.917 for frond, 0.931 for root, and 0.939 for the combined model, with errors concentrated at adjacent class boundaries. These results show that VNA-based dielectric measurement combined with simple feature engineering and per-part models can detect and classify BSR severity accurately and reproducibly, providing an ex vivo proof-of-concept baseline to guide future field validation and development toward more practical measurement approaches.