Prototype of a Edge-AI Based Bird Repellent System  Using Yolov8

Gustia Fernando; Ilmiyati Rahmi Jasril; Almasri Almasri; Rido Putra

doi:10.24036/voteteknika.v14i1.137734

Prototype of a Edge-AI Based Bird Repellent System Using Yolov8

Gustia Fernando⁽¹⁾, Ilmiyati Rahmi Jasril⁽²⁾, Almasri Almasri⁽³⁾, Rido Putra⁽⁴⁾,
(1) Universitas Negeri Padang

Indonesia
(2) Universitas Negeri Padang

Indonesia
(3) Universitas negeri padang

Indonesia
(4)

Indonesia

Corresponding Author

DOI : https://doi.org/10.24036/voteteknika.v14i1.137734

Full Text:

Language : en

Abstract

Bird pest control in agricultural areas generally employs conventional methods that are susceptible to habituation processes and lack operational efficiency. Cloud-based computer vision implementations suffer from network latency vulnerabilities , while the utilization of standard Single Board Computers (SBCs) frequently triggers computational bottlenecks when processing deep learning algorithms. Therefore, this study proposes the design of a smart pest repellent prototype based on Edge Artificial Intelligence (Edge-AI) exploiting the Neural Processing Unit (NPU) on the Orange Pi 5 Pro platform. The object detection model was developed utilizing the YOLOv8nn architecture, trained on a dataset of 40,877 images, and executed locally. The system implements stochastic stimulus control logic through the integration of a pan-tilt green laser actuator and predator audio emissions to prevent habituation processes in pests. System performance testing results indicate that NPU acceleration is capable of yielding a stable inference throughput at 20 Frames Per Second (FPS). The model's performance evaluation achieved a Mean Average Precision (mAP@50) of 0.70037 , with an optimal visual detection range effectiveness of 4 meters. Under peak computational and actuation conditions (full load), the instrument recorded a power consumption of 21.09 W, equivalent to an estimated operational viability of 2.8 hours utilizing a 7,200 mAh battery supply. In conclusion, the proposed Edge-AI architecture proves robust in mitigating embedded computational limitations, yielding a control instrument that is responsive, accurate, and highly reliable for real-world environmental applications.

Keywords – Computer Vision, Edge-AI, NPU, Orange Pi 5 Pro, YOLOv8n.

References

J. L. Mela and C. G. Sánchez, “Yolo-based power-efficient object detection on edge devices for USVs,” J. Real. Time. Image Process., vol. 22, no. 3, Jun. 2025, doi: 10.1007/s11554-025-01682-2.

W. I. Zen and I. R. Jasril, “Pengembangan Sistem Penyiangan Gulma Padi Berbasis IoT Menggunakan ESP32 untuk Meningkatkan Produktivitas Pertanian,” vol. 2, no. 2, 2024, doi: 10.24036/elektif.v2i2.51.

R. Yauri, E. Campos, R. Yalico, and V. Gamero, “Development of an Electronic Bird Repellent System using Sound Emission,” WSEAS Transactions on Systems and Control, vol. 18, pp. 136–143, 2023, doi: 10.37394/23203.2023.18.14.

P. Dhilleswararao, S. Boppu, M. S. Manikandan, and L. R. Cenkeramaddi, “Efficient Hardware Architectures for Accelerating Deep Neural Networks: Survey,” 2022, Institute of Electrical and Electronics Engineers Inc. doi: 10.1109/ACCESS.2022.3229767.

D. K. Alqahtani, A. Cheema, and A. N. Toosi, “Benchmarking Deep Learning Models for Object Detection on Edge Computing Devices,” Sep. 2024, [Online]. Available: http://arxiv.org/abs/2409.16808

S. J. Poornashree, P. V. Joshi, K. M. Sudharshan, and T. M. George, “A Hybrid AI Pipeline for Real-Time Aerial Video Analytics on Resource-Limited Edge Devices with Performance Profiling,” Engineering, Technology and Applied Science Research, vol. 15, no. 6, pp. 29574–29579, Dec. 2025, doi: 10.48084/etasr.14396.

J. L. Mela and C. G. Sánchez, “Correction: Yolo-based power-efficient object detection on edge devices for USVs (Journal of Real-Time Image Processing, (2025), 22, 3, (108), 10.1007/s11554-025-01682-2),” Dec. 01, 2025, Springer Nature. doi: 10.1007/s11554-025-01776-x.

E. Manor and S. Greenberg, “Custom Hardware Inference Accelerator for TensorFlow Lite for Microcontrollers,” IEEE Access, vol. 10, pp. 73484–73493, 2022, doi: 10.1109/ACCESS.2022.3189776.

A. Garcia-Perez, R. Miñón, A. I. Torre-Bastida, and E. Zulueta-Guerrero, “Analysing Edge Computing Devices for the Deployment of Embedded AI,” Sensors, vol. 23, no. 23, Dec. 2023, doi: 10.3390/s23239495.

C. Khelfa, H. Drias, I. Khennak, and K. Elleithy, “GPU-Accelerated Slime Mould Algorithm for Urgent Transportation in Disaster Response: A COVID-19 Application,” 2025.

Ian. Sommerville, Software engineering. Pearson, 2016.

Sparrow, “sparrow Computer vision Dataset.” Accessed: Feb. 23, 2026. [Online]. Available: https://universe.roboflow.com/sparrow-0jml6/sparrow-qnulu

Sparrow Bird, “sparrow Bird Computer Vision Dataset .” Accessed: Feb. 23, 2026. [Online]. Available: https://universe.roboflow.com/sparrow-bird/sparrow-bird-g6dxj

Universe, “Sparrow Object Detection Model by Project.” Accessed: Feb. 23, 2026. [Online]. Available: https://universe.roboflow.com/project-57vuf/sparrow-dgtbl

Ultralytics, “Explore Ultralytics YOLOv8n - Ultralytics YOLO Docs.” Accessed: Feb. 25, 2026. [Online]. Available: https://docs.ultralytics.com/models/YOLOv8n/

H. Rezatofighi, N. Tsoi, J. Gwak, A. Sadeghian, I. Reid, and S. Savarese, “Generalized Intersection over Union: A Metric and A Loss for Bounding Box Regression,” Apr. 2019, [Online]. Available: http://arxiv.org/abs/1902.09630

X. Tu et al., “Unveiling Energy Efficiency in Deep Learning: Measurement, Prediction, and Scoring Across Edge Devices,” in Proceedings - 2023 IEEE/ACM Symposium on Edge Computing, SEC 2023, Institute of Electrical and Electronics Engineers Inc., 2023, pp. 80–93. doi: 10.1145/3583740.3628442.

N. Rai et al., “Applications of deep learning in precision weed management: A review,” Comput. Electron. Agric., vol. 206, Mar. 2023, doi: 10.1016/j.compag.2023.107698.

D. Andrea, “Battery Management Systems for Large Lithium-Ion Battery Packs,” 2010.

Orange Pi Ltc, “Orange Pi 5 Pro.” Accessed: Feb. 23, 2026. [Online]. Available: http://www.orangepi.org/html/hardWare/computerAndMicrocontrollers/details/Orange-Pi-5-Pro.html

Lenovo, “Lenovo LOQ 15IAX9.” Accessed: Feb. 23, 2026. [Online]. Available: https://www.lenovo.com/id/id/p/laptops/loq-laptops/lenovo-loq-15iax9/len101q0006?orgRef=https%253A%252F%252Fwww.google.com%252F&srsltid=AfmBOop7e-mwecEMrNhfbqxQGq7za9pISjG1ddnffTuLQ44348ZG_-tD

“MG995 Datasheet(PDF) - List of Unclassifed Manufacturers.” Accessed: Feb. 23, 2026. [Online]. Available: https://www.alldatasheet.com/datasheet-pdf/pdf/1132435/ETC2/MG995.html

“balenaEtcher - Flash OS images to SD cards & USB drives.” Accessed: Feb. 26, 2026. [Online]. Available: https://etcher.balena.io/

Armbian, “Orange Pi - Armbian,” Armbian. Accessed: Dec. 22, 2025. [Online]. Available: https://www.armbian.com/orange-pi-3/

A. Ramaditiya, S. Rahmatia, A. Munawar, and O. N. Samijayani, “Implementation chatbot whatsapp using python programming for broadcast and reply message automatically,” in Proceeding - 2021 International Symposium on Electronics and Smart Devices: Intelligent Systems for Present and Future Challenges, ISESD 2021, Institute of Electrical and Electronics Engineers Inc., Jun. 2021. doi: 10.1109/ISESD53023.2021.9501523.

Python, “Python.org Documentation.” Accessed: Feb. 26, 2026. [Online]. Available: https://www.python.org/doc/

“Roboflow: Computer vision tools for developers and enterprises.” Accessed: Feb. 26, 2026. [Online]. Available: https://roboflow.com/

M. Qian et al., “Real time wire rope detection method based on Rockchip RK3588,” Sci. Rep., vol. 15, no. 1, Dec. 2025, doi: 10.1038/s41598-025-16043-z.

Q. Chen et al., “An experimental study of acoustic bird repellents for reducing bird encroachment in pear orchards,” Front. Plant Sci., vol. 15, 2024, doi: 10.3389/fpls.2024.1365275.