View Proposal

---

Proposer

Chee Een Yap

Title

Development of a Low-Cost Autonomous Delivery Vehicle with Object Shape Recognition Using Zero-Shot Learning

Goal

This study aims to develop a low-cost intelligent autonomous delivery vehicle that integrates computer vision and Zero-Shot Learning for object shape recognition and automated logistics management. The system identifies the shape of delivery items using a recognition camera, unlocks the corresponding storage compartment for item placement, and automatically relocks the compartment after 30 seconds to ensure secure transportation and efficient closed-loop logistics operations.

Description

This project proposes the development of an Autonomous Delivery Vehicle that integrates computer vision on Raspberry Pi with Zero-Shot Learning (ZSL) to automate item storage and delivery processes. The system uses a low-cost camera with a Raspberry Pi to recognize the shape of delivery items, including previously unseen objects, without requiring training data for every category. Based on the recognition result, the system automatically unlocks the corresponding storage compartment for item placement. After 30 seconds, the electromagnetic lock automatically re-engages to secure the compartment. By performing all processing locally on the vehicle, the proposed solution minimizes reliance on cloud connectivity while providing a cost-effective, secure, and efficient logistics platform.

Resources

Background

Url

Difficulty Level

Easy

Ethical Approval

None

Number Of Students

1

Supervisor

Chee Een Yap

Keywords

raspberry pi, zero-shot learning (zsl), object recognition

Degrees

Bachelor of Science in Computing Science