Optimizing IoT-Enabled Logistics Systems Using Lightweight Neural Networks for Real-Time Predictive Scheduling
Abstract
advanced AI use. This paper presents a lightweight CNN framework based on MobileNetV2 for predictive scheduling in IoT logistics. Using
GPS, traffic, and delivery status data, it achieves 87.5% accuracy, 32% lower latency than heuristic methods, and 64% less memory than typical CNNs. Developed from the author’s expertise in IoT, AI, and logistics, the framework supports last-mile delivery, smart city logistics, and
industrial IoT. It validates the author’s master’s curriculum in AI, IoT, robotics, and cybersecurity at Hong Kong universities..
Keywords
Full Text:
PDFReferences
[1]Howard, A. G., et al. (2017). MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications. arXiv preprint
arXiv:1704.04861.
[2]Sandler, M., et al. (2018). MobileNetV2: Inverted Residuals and Linear Bottlenecks. CVPR, 4510–4520.
[3]Tan, M., & Le, Q. V. (2019). EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks. ICML, 6105–6114.
[4]Li, X., et al. (2019). IoT-Based Supply Chain Management. IEEE Internet of Things Journal, 6(3), 4567–4580.
[5]Wang, Y., et al. (2020). Deep Learning for Delivery Time Prediction. Journal of Network and Computer Applications, 165, 102632.
[6]Chen, Z., et al. (2021). Lightweight AI Models for IoT-Based Smart Agriculture. Sensors, 21(12), 4089.
[7]Lee, J., et al. (2020). Security Challenges in IoT-Based Logistics Systems. IEEE Transactions on Information Forensics and Security, 15, 1234–1245.
[8]Zhang, H., et al. (2022). Edge Computing for IoT-Enabled Logistics. Journal of IoT, 4(2), 89–102.
[9]MQTT Version 5.0 Specification. (2019). OASIS Standard.
DOI: http://dx.doi.org/10.18686/ede.v4i2.14172
Refbacks
- There are currently no refbacks.