This paper proposes HaloBus, an innovative, edge-computing solution designed to mitigate this risk by detecting student boarding and exiting in real time using lightweight AI based methods. A persistent challenge in elementary school transportation is the issue of missing students after they exit their buses, which disproportionately impacts low-income households. Current safety systems place the burden of implementation on individual households, often requiring independent methods. Common methods include applications on a personal device or a small tracker. However, not everyone can afford these options, and ensuring child safety is a primary concern for parents and caregivers. That is why HaloBus was invented. The system employs YOLOv5us—an Ultralytics-enhanced, anchor-free, split-head architecture that offers a superior accuracy speed trade-off. By providing real-time, on-device alerts, HaloBus enables immediate intervention to prevent a student from being left behind, thereby shifting the focus from reactive post-incident response to proactive safety. Trained on over 70,000 labeled and unlabeled images, the model can accurately detect multiple students simultaneously, significantly reducing false positives. In real-world deployment, the model sustained 30 frames per second on the Raspberry Pi and achieved detection confidence levels exceeding 75% even when subjects wore sunglasses or hoodies. With opt-in participation for each family, HaloBus effectively balances detection efficiency and privacy protection. Overall, HaloBus offers a low-cost, scalable, and ethically conscious approach to enhancing school-bus safety by delivering reliable, on-device boarding and exit detection for multiple students in varied real-world conditions.