MobilePlan - Optimization-powered mobile manipulation task planning for agriculture pick-and-carry tasks
The proposed TestBed offers an integrated platform for validating multi-robot autonomous crop collection in realistic agricultural logistics scenarios. Its core technology focus is the coordination of heterogeneous mobile manipulators through an optimization-based global planner, where the Crop Collection Problem is formulated as a Flexible Multi-Depot Capacitated Vehicle Routing Pickup Problem (FMDCVRP-P). This approach enables robots to dynamically start and end at different depots, reducing unnecessary travel and minimizing overall mission makespan.
The TestBed combines Mixed-Integer Linear Programming (MILP) optimization with a full ROS2–ROS1 simulation and deployment pipeline, including Gazebo-based multi-robot environments, task scheduling, MoveIt-enabled manipulation, and autonomous navigation using AMCL and TEB planners integrated for the RBKAIROS+ Mobile Manipulator.
This TestBed is highly relevant to DeepTech and Industry 5.0, supporting scalable, intelligent, and human-centered automation for next-generation sustainable agriculture and autonomous logistics.
| Name of Principal Investigator: | Kosta Jovanović |
| Position / institutional role: | Associate professor |
| Info Email: | kostaj@etf.rs |
| ORCID persistent identifier (PID): | 0000-0002-9029-4465 |
| Name of Host Organization | University of Belgrade – School of Electrical Engineering (ETF) |
| Department or Lab | Department of Signals and Systems, ETF Robotics lab |
| Name of Building | Palace of Science |
| Physical Address | Kralja Milana 11, 11000 Beograd, Serbia |
| Website Links | https://robot.etf.bg.ac.rs |
| Institutional contact name | Nikola Knezevic |
| Institutional contact email | knezevic@etf.rs |
| Non-academic stakeholders |
| Industrial Partners; Startups; Professional Associations; SMEs
Government Bodies; Community |
| Academic stakeholders |
| Undergraduate students; PhD students; MSc students; Researchers |
| Other types of stakeholders |
| agri-tech integrators, logistics operators, farm cooperatives |
| Application case: | Short description: |
| Multi-Robot Crop Collection Route Optimization (PoC) | validation of optimization-based task allocation and routing for fleets of mobile manipulators performing post-harvest crop bin pickup and delivery while minimizing makespan and respecting payload and depot constraints |
| Autonomous Navigation and Obstacle Avoidance in Unstructured Environments | testing robust navigation of RB-KAIROS platforms using AMCL and TEB in agricultural or industrial spaces with sparse features and non-uniform obstacles |
| Mobile Manipulation Pick-and-Carry Experiments | integration of manipulation planning with autonomous base positioning to perform crop-bin pickup, transport, and delivery missions in simulation and on the real platform |
| Educational Use in Robotics and Autonomous Systems Courses | supports coursework and student projects in ROS, optimization, mobile manipulation, and Industry 5.0 automation |
list of hardware components with their brief descriptions:
1. RB-KAIROS Mobile Manipulator Platform (Robotnik) – omnidirectional mobile base with integrated manipulation support, suited for autonomous navigation in industrial logistics and field robotics. Link: https://robotnik.eu/products/mobile-robots/rb-kairos-2/
2. Franka Emika Research 3 (FR3) collaborative manipulator – 7 DoF torque-controlled robotic arm, payload 3 kg, reach 855 mm, suitable for safe manipulation experiments. Link: https://franka.de/franka-research-3
3. Robotics workstation / compute server – multi-core CPU and GPU-enabled system for MILP optimization, ROS middleware, Gazebo simulation, and motion planning.
list of software components with their brief descriptions:
1. Ubuntu Linux 20.04/22.04 LTS – primary OS for ROS-based robotics development. Link: https://ubuntu.com/download
2. ROS1 Noetic and ROS2 Humble/Iron – middleware for robot control, navigation, manipulation, and multi-robot coordination. Link: https://www.ros.org/
3. MoveIt Motion Planning Framework – arm manipulation planning and execution. Link: https://moveit.ros.org/
4. ROS Navigation Stack with AMCL and TEB local planner – autonomous navigation, localization, and obstacle avoidance. Links: https://wiki.ros.org/navigation and https://wiki.ros.org/teb_local_planner
5. Gurobi Optimizer – MILP solver for the Flexible Multi-Depot Capacitated VRP Pickup formulation. Link: https://www.gurobi.com/
6. Continuous development platform for RB-KAIROS robot control and planning. Link: https://github.com/etfrobotics/rbkairos_etf_services