Simulation Environment for User-Robot Voice Interaction in Product Handling and Packaging Systems

Advances in artificial intelligence, robotics, and human-machine interaction have permeated industrial processes, making them smarter, more flexible, and more autonomous. Voice interaction is a natural medium that has become a key factor in improving production systems, applicable to processes such as order picking. In this context, this article aimed to develop an intelligent voice-controlled packaging system that integrates natural language processing, computer vision, and robotic product handling, which was then evaluated in a virtual simulation environment. The methodology consisted of designing a simulated environment, developing and integrating voice recognition and synthesis algorithms into an automated workflow that allows the selection and manipulation of finished products to prepare orders. The system was implemented using a robotic platform, a conveyor belt, and a finished goods storage area. An interface was created to facilitate control of the packaging process through a chatbot capable of understanding voice commands and responding to the user utilizing voice recognition and synthesis algorithms. The system's accuracy and robustness were evaluated using 20 commands, analyzed with recordings generated for 13 user profiles with different accents, ages, and vocal characteristics. The results showed an average accuracy of 85.7% in command transcription, with robust performance against voice variations, although the lowest accuracy was observed when using children's voices. Additionally, the packaging and sorting system was validated in the virtual environment, demonstrating its efficient operation in managing box space and available shelf inventory. Thanks to the results obtained, it is possible to conclude that the designed system allows for almost natural and flexible interaction through voice commands, effectively integrating language recognition and robotic manipulation in a simulated environment, which demonstrates its potential application in supply chain automation and Industry 5.0.