Izvorni znanstveni članak

https://doi.org/10.17559/TV-20250320002497

Affordable Collision Detection for a 3D-Printed Robotic Arm

Almir Osmanović orcid.org/0000-0002-6839-1234 ; University of Tuzla, Faculty of Mechanical Engineering, Univerzitetska 4, 75000 Tuzla, Bosnia and Herzegovina *
Malik Čabaravdić orcid.org/0009-0008-0302-8216 ; University of Zenica, Faculty of Mechanical Engineering, Fakultetska 3, Zenica 72000, Bosnia and Herzegovina
Amer Sarajlić orcid.org/0009-0009-1626-4661 ; University of Zenica, Faculty of Mechanical Engineering, Fakultetska 3, Zenica 72000, Bosnia and Herzegovina
Jasmin Halilović orcid.org/0009-0005-5265-4561 ; University of Tuzla, Faculty of Mechanical Engineering Tuzla
Kenan Varda ; University of Zenica, Faculty of Mechanical Engineering, Fakultetska 3, Zenica 72000, Bosnia and Herzegovina

* Dopisni autor.

Sažetak

Collision detection and avoidance are crucial for safe and efficient robotic manipulation. This paper presents an affordable collision detection system for a 3D-printed robotic arm using a Kinect camera integrated with the Robot Operating System (ROS). The proposed approach enables real-time environment perception and collision-free trajectory generation. Experimental results show a significant improvement in collision avoidance efficiency, achieving a 40% reduction in path length while maintaining real-time performance. The computational overhead introduced by collision detection is minimal, with a 12 ms delay per frame, which can be optimized further. It is demonstrating the system’s ability to effectively handle dynamic environments, such as human obstacles. This work highlights the potential of low-cost 3D sensing solutions for robotics applications, making them more accessible and scalable.

Ključne riječi

collision detection; kinect; mechatronic; robotic; robotic arm; ROS; trajectory optimization

Hrčak ID:

337707

URI

https://hrcak.srce.hr/337707

Datum izdavanja:

31.10.2025.

Posjeta: 668 *