*Result*: INTELLIGENT CONTROL OF A ROBOTIC ARM FOR IMPROVED PICK-AND-PLACE PERFORMANCE.

*Traditional pick-and-place robotic arms generally rely on fixed positioning and analogue control mechanisms, which constrain their flexibility and lower the accuracy of object recognition in dynamic settings. These designs, which often utilise potentiometer-based inputs and basic visual systems, typically reach only 70-85% accuracy due to limited field of view, inadequate adaptation to lighting changes, and immobility. To address these issues, this research presents a mobile intelligent pick-and-place robotic arm featuring an ESP32-CAM and a classification system built on TinyML (CNN). By acquiring images of objects from multiple perspectives and evaluating RGB data using a weighted voting approach, the robot greatly minimizes errors caused by shadows and lighting variations. Its mobile platform improves alignment with targets and enhances handling precision, achieving a 95% classification accuracy--an increase of 10-25% over traditional models. Furthermore, integration with an Arduino Nano and HC-05 Bluetooth module facilitates remote control through an Android application, making the system effective for deployment in risky or difficult-to-access areas. The proposed solution offers potential for use in automated logistics, sorting systems, and handling dangerous materials. Future enhancements will explore the use of deep learning and reinforcement learning to support advanced recognition and adaptive control without manual code adjustments. [ABSTRACT FROM AUTHOR]

Copyright of Annals of the Faculty of Engineering Hunedoara - International Journal of Engineering is the property of University Politehnica Timisoara, Faculty of Engineering Hunedoara and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)*