ORIGINAL ARTICLE
Vision-Guided Cooperative Manipulation Framework for
Synchronized Grasping and Transport of Cylindrical Objects Using
Dual Mobile Manipulators
More details
Hide details
1
Department of Control and Instrumentation Engineering, King Fahd University of Petroleum and Minerals, Saudi Arabia
Submission date: 2026-04-15
Final revision date: 2026-05-08
Acceptance date: 2026-08-23
Publication date: 2026-11-16
Corresponding author
Muhammad Faizan Mysorewala
Department of Control and Instrumentation Engineering, King Fahd University of Petroleum and Minerals, KFUPM, 31261, Dhahran, Saudi Arabia
Journal of Undergraduate Research International 2026;2(3B):74-81
KEYWORDS
TOPICS
ABSTRACT
Despite advances in cooperative mobile manipulation, reports on low-cost frameworks that enable dual-robot grasping and transport
without specialized hardware remain limited. This study introduces and experimentally validates a vision-guided, synchronized
cooperative manipulation framework for two DJI RoboMaster EP mobile manipulators performing a cylindrical object pick-andplace
task in warehouse-relevant settings. The key contribution of this work is an event-driven synchronization mechanism for
dual-robot cooperative manipulation, which enables coordinated grasping and transport without direct robot-to-robot wireless
messaging, depth sensors, external motion-capture infrastructure, or pre-built 3D maps. The manipulation pipeline includes visual
lateral alignment, visual forward approach, grasp execution, synchronized lifting, cooperative transport, and coordinated release.
Visual guidance is achieved using color-based HSV detection and proportional control, providing a lightweight perception and
control solution for cooperative object handling. The framework is evaluated through fixed-position repeatability and position generalization
studies. In a five-trial repeatability study, the system achieves a 100% task success rate with a mean completion time of
65.26 ±1.65 s and coefficient of variation of 2.5%. In a six-configuration generalization study, the system achieves 100% success,
with task completion times ranging from 50.8 s to 86.5 s. Across all eleven trials, the proposed scheme maintains safe and correct
task progression. Grip synchronization latency increases with asymmetry in the robots’ approach distances, with symmetric and
asymmetric configurations producing latencies ≤ 6 s and ≥ 22 s, respectively. Overall, event-driven synchronization offers a practical,
lightweight coordination mechanism for cooperative dual-robot manipulation under controlled laboratory conditions, with
minimal sensing and communication requirements.