Integrated Hardware and Software Architecture for Industrial AGV with Manual Override Capability
Pietro Iob, Mauro Schiavo, Angelo Cenedese
TL;DR
The paper tackles automating hazardous industrial transport by converting a traditional human-operated vehicle into a fully autonomous AGV, with a mandatory manual override for safety. It proposes an integrated hardware-software platform built around a drive-by-wire system, sensor-rich perception, and a modular software stack (Sensing, Planning, Control) orchestrated by a system supervisor. A dual-controller hardware architecture (ADPU and PLC) supports localization, navigation, and safety supervision, while a Finite State Machine enables seamless switching between manual and autonomous modes. Preliminary experiments demonstrate bounded response times and safe coexistence of human and autonomous control, highlighting practical benefits for safety and productivity in industrial settings such as the primary aluminum industry.
Abstract
This paper presents a study on transforming a traditional human-operated vehicle into a fully autonomous device. By leveraging previous research and state-of-the-art technologies, the study addresses autonomy, safety, and operational efficiency in industrial environments. Motivated by the demand for automation in hazardous and complex industries, the autonomous system integrates sensors, actuators, advanced control algorithms, and communication systems to enhance safety, streamline processes, and improve productivity. The paper covers system requirements, hardware architecture, software framework and preliminary results. This research offers insights into designing and implementing autonomous capabilities in human-operated vehicles, with implications for improving safety and efficiency in various industrial sectors.