AI-Powered Real-Time Package Drop Warning System

Overview

This project focused on the design and testing of an integrated electronic system capable of detecting, classifying, and wirelessly reporting damaging package handling events in real time. A self-contained device was built using an Arduino Nano 33 BLE Sense microcontroller with built-in sensors, a SIM7600G‑H 4G LTE module for cellular communication and GPS, and a custom power management system connected to a Li-ion battery supporting 2A peak currents. Over ten hours of accelerometer data across six package handling classes (idle, driving, walking, drop, throw, and vigorous shake) were collected and used to train a neural network classifier on Edge Impulse, which was then deployed to the Arduino for on‑device real‑time inference. 22 instrumented shipments totaling 48 alerts were conducted to validate the device performance. The system achieved 99.6% recall in detecting damaging events with 99.9% overall accuracy distinguishing normal from alert conditions, and battery runtime averaged 106 hours. Analysis of alert data revealed that 78% of damaging events occurred at sorting facilities and during last‑mile delivery, with drops three times more frequent than throws or shakes at hubs. This work demonstrates that a low‑cost, battery‑powered electronic system with on‑device AI can provide reliable real‑time monitoring of package handling, offering actionable data to improve supply chain accountability and protect sensitive cargo.

Competition history

• CSEF 2026 Electronics & Electromagnetics (Senior Division) · Entry S-10-09

Related projects

Closest projects by meaning, across every fair and year in the corpus.

Browse more like this

• All Physics and Astronomy projects
• CSEF Physics and Astronomy projects
• Every CSEF 2026 project