Robotics and Automation
[Open soon]
Work description
The scholarship recipient will participate in the different stages of the research cycle, from problem analysis and algorithm development to its implementation and experimental evaluation on a real robotic platform: 1. Study of the state of the art and definition of requirements A critical analysis of existing approaches for robotic box handling, humanoid robotics, bimanual grasping, computer vision, and automation of return processes will be carried out: • survey of the main techniques and platforms available; • characterization of the boxes and logistical scenarios considered; • identification of safety, reach, load, balance, and workspace constraints; • definition of handling tasks, test scenarios, and evaluation metrics. 2. Preparation and integration of the humanoid platform The configuration of the robotic platform and the components necessary for the execution of the proposed tasks will be carried out, including sensors, cameras, communication interfaces, and control modules: • configuration of the development and simulation environment; • calibration of the robot's arms, hands, and sensors; • calibration between the robot and the vision systems; • Integration of RGB and/or Lidar cameras; • Development of communication interfaces between perception, planning, and control modules; • Definition of safety, supervision, and operation shutdown mechanisms. 3. Development of the perception system Methods will be developed to allow the robot to detect boxes, estimate their position and orientation, and identify suitable points for their manipulation. Classical approaches based on image processing and point clouds, as well as segmentation and classification methods based on artificial intelligence, will be studied. The system should be able to handle variations in: • Dimension, geometry, and appearance of the boxes; • Position and orientation; • Lighting; • Partial occlusions; • Arrangement of objects in the workspace. 4. Development of bimanual gripping strategies Strategies will be investigated to allow the robot to use both arms in a coordinated manner to safely grasp, lift, transport, and place boxes: • Definition of suitable gripping points; • Automatic generation of arm and hand configurations; • Fine visual adjustment of the gripping pose; • Bimanual coordination during lifting and transport; • Adaptation of the strategy to the dimensions and mass of the boxes; • Detection of unstable grips and recovery in case of failures. 5. Planning and control of handling tasks Algorithms will be developed to autonomously perform pick-and-place operations at defined points in the sorting zone: • Planning collision-free trajectories; • Coordination between arms, torso, and robot posture; • Control of approach, gripping, lifting, and placement of boxes; • Sequencing of the different actions of the task; • Real-time monitoring of execution; • Implementation of recovery mechanisms in case of perception, planning, or handling errors. 6. Integration of visual identification and inspection resources The possibility of using the robot's vision system to support the sorting process and the digital recording of returns will be studied. Functionalities such as the following may be implemented: • Reading barcodes and QR codes; • Recognition of fiducial markers; • Identification of the box or product; • Verification of the presence and arrangement of items; • Visual classification or segmentation using artificial intelligence; • Association of inspection results with the digital process management system. 7. Experimental Validation and Performance Evaluation The developed modules will initially be tested in a controlled environment and subsequently evaluated in scenarios representative of the LOGIX project's return flows. The evaluation will consider, among others, the following indicators: • Success rate in detecting and locating boxes; • Success rate of grasping and placing; • Number of boxes handled per unit of time; • Average execution time of each cycle; • Placement accuracy; • Occurrence of collisions or unsafe situations; • Robustness in the face of variations in boxes and the environment; • Feasibility of visual inspection and reintroduction into the logistics flow. 8. Documentation, Analysis, and Dissemination of Results Documentation of the developed methodologies and critical analysis of experimental results: • preparation of technical documentation for the software and experimental procedures; • creation of test protocols and datasets; • preparation of system demonstrations; • participation in the preparation of the performance report for the humanoid robot control module; • contribution to scientific publications, technical communications, or other actions to disseminate the results.
Academic Qualifications
Master's degree in Electrical Engineering or related fields.
Minimum profile required
The candidate must meet the following requirements:• Completed Master's degree in Electrical and Computer Engineering or related fields.• Be enrolled in a doctoral program in Electrical and Computer Engineering or related fields, or in a non-degree-granting program, under the terms applicable to the scholarship regulations.• Proven experience in developing R&D projects with industry.• Fundamental knowledge of robotics, including kinematics, motion planning, control, or robotic manipulation.• Experience in the development, integration, or validation of robotic solutions and/or automated systems in a laboratory or industrial environment.• Knowledge of C++ and Python programming.• Knowledge and experience with the ROS/ROS2 framework.
Preference factors
The following elements will be valued: • Knowledge and experience with OpenCV and PCL libraries. • Experience with RGB-D cameras and point cloud processing. • Experience with robotic simulators, such as Gazebo or Isaac Sim. • Participation in robotics competitions. • Fluency in spoken and written English.
Application Period
Since 28 Jul 2026 to 10 Aug 2026
[Open soon]
Centre
Robotics in Industry and Intelligent Systems