Robotics AI Research Fellow: Sensor Fusion & 3D Perception

Robotics AI Research Fellow: Sensor Fusion & 3D Perception
Coimbra
Coimbra, Coimbra District, Portugal

Organisation/Company Instituto Politécnico de Coimbra Department DGRH Research Field Other Researcher Profile First Stage Researcher (R1) Positions Master Positions Country Portugal Application Deadline 9 Dec 2025 - 23:59 (Europe/Lisbon) Type of Contract Temporary Job Status Full-time Hours Per Week 35 Offer Starting Date 16 Jan 2026 Is the job funded through the EU Research Framework Programme? Not funded by a EU programme Reference Number PRBI/32/2025 Is the Job related to staff position within a Research Infrastructure? No

Offer Description

The Polytechnic University of Coimbra opens a call for one research fellowships (BI), in the framework of the the CRIARTE project – Ref. : COMPETE2030-FEDER-01192200 - 17402,
- financed by the Portugal 2030 program – European Regional Development Fund (ERDF) through the Thematic Program Innovation and Digital Transition – Compete 2030, under the following conditions:

The scholarship aims to support the implementation and development of the following activities:

• Sensor Integration and Robotic Perception
• Expertise in sensor fusion and sensor calibration;
• Experience with Li
  DAR, RGB-D/multimodal cameras, GNSS, IMU, UWB sensors, and Io
  T devices;
• Proficiency in ROS/ROS2 (Robot Operating System) and middleware for data integration;
• Knowledge of temporal and spatial synchronization of sensors and extrinsic/intrinsic calibration.
• Computer Vision and 3D Pose Estimation
• Strong background in computer vision and machine learning applied to pose estimation and visual servoing;
• Experience with Open
  CV, PCL (Point Cloud Library), Py
  Torch/Tensor
  Flow, and 3D frameworks such as COLMAP and Open3D;
• Ability to develop algorithms for object detection and tracking, 3D reconstruction, and SLAM.
• Advanced Control and Intelligent Robotics
• Solid knowledge of classical and modern control theory (PID, adaptive, robust, and fractional control);
• Experience with fuzzy logic and AI-based control (reinforcement learning,
  - fuzzy systems);
• Skills in modelling and simulation of dynamic systems (MATLAB/Simulink, Gazebo);
• Familiarity with robotic manipulators and manipulator kinematics/dynamics.
• Digital Twin Development and Simulation
• Experience in digital modelling of robotic systems and
  - physical integration;
• Knowledge of digital twin tools such as Unity, Unreal Engine, Siemens NX, and MATLAB Simscape;
• Ability to represent complex kinematic and dynamic models, including noise and sensor uncertainties.
• Applied Artificial Intelligence and Data Analysis
• Proficiency in multivariate analysis techniques (PCA, regression, clustering) and predictive risk models;
• Experience with artificial neural networks and fuzzy systems for KPI aggregation;
• Knowledge in data engineering, Python, and
  - time data analysis.

Generic Scientific Area: Electrical Engineering or Computer Engineering

Specific Scientific Area: Automation and Robotics

Master’s degree in Electrical Engineering or related fields; enrolled in a Ph
D program or a
- degree course.

Experience in:

a) ROS, programming in C++ and Python;

b) Modelling and control.

Work Plan:

The work plan is divided into five phases:

Phase 1 – Development of Software for Advanced Sensor Integration. Objective: Create a sensor integration infrastructure combining data from multiple sensors and human inputs. Activities: Survey and characterization of robot sensors (Li
DAR, multimodal cameras, GNSS, IMU, etc. ) and external sensors (CCTV, UWB, Io
T); development of data acquisition, filtering, and preprocessing pipelines; implementation of sensor fusion algorithms for temporal and spatial integration; integration of human inputs from Augmented Reality (A3) devices and other collaborative sensors; validation tests in simulated and real environments.

Phase 2 – Design of AI Architecture for 3D Pose Estimation. Objective: Develop an AI architecture capable of estimating the 3D position and orientation of objects with robustness and adaptability. Activities: Implementation of visual servoing and 3D pose estimation algorithms; integration of adaptive control techniques, including fractional PID and fuzzy logic; development of modules to handle uncertainty and noise in visual data; performance testing under various construction scenarios and environmental conditions; dynamic adjustment of control parameters for maximum precision and robustness.

Phase 3 – Development of Control Architecture for Sandwich Panel Manipulation. Objective: Create an advanced and collaborative control system for precise and safe manipulation of panels. Activities: Modelling of panels and manipulation devices based on environmental sensors; implementation of visual servoing for dynamic control of panel position and orientation; application of advanced control methods (PID, fuzzy logic, fractional control) to reduce errors; integration of perception and control systems to adapt to obstacles and environmental variability; experimental validation of manipulation accuracy, safety, and efficiency.

Phase 4 – Development of the Robot Digital Twin. Objective: Create a detailed digital model of the robotic platforms for simulation, analysis, and optimization. Activities: Nonlinear kinematic and dynamic modelling of mechanical components (rotary and linear joints); integration of accurate sensor models (Li
DAR, multimodal cameras) including associated errors; simulation of robot behavior under various operational conditions; optimization tests of the
- action architecture; predictive analysis of potential failures and continuous performance improvement.

Phase 5 – Integration of Key Performance Indicators (KPIs). Objective: Create a risk assessment system based on data from assets, humans, and robots to support
- making. Activities: Definition of a multidimensional matrix of KPIs and risk levels; development of indicator aggregation techniques (weighted average, direct sum, PCA, multivariate regression, fuzzy logic, ANN); implementation of dynamic risk score adjustment based on the actual execution state; simulation and analysis of risk evolution during project execution; optimization of resource allocation and
- making support using integrated KPIs.

The scholarship aims to support the implementation and development of the following activities:

• Sensor Integration and Robotic Perception
• Expertise in sensor fusion and sensor calibration;
• Experience with Li
  DAR, RGB-D/multimodal cameras, GNSS, IMU, UWB sensors, and Io
  T devices;
• Proficiency in ROS/ROS2 (Robot Operating System) and middleware for data integration;
• Knowledge of temporal and spatial synchronization of sensors and extrinsic/intrinsic calibration.
• Computer Vision and 3D Pose Estimation
• Strong background in computer vision and machine learning applied to pose estimation and visual servoing;
• Experience with Open
  CV, PCL (Point Cloud Library), Py
  Torch/Tensor
  Flow, and 3D frameworks such as COLMAP and Open3D;
• Ability to develop algorithms for object detection and tracking, 3D reconstruction, and SLAM.
• Advanced Control and Intelligent Robotics
• Solid knowledge of classical and modern control theory (PID, adaptive, robust, and fractional control);
• Experience with fuzzy logic and AI-based control (reinforcement learning,
  - fuzzy systems);
• Skills in modelling and simulation of dynamic systems (MATLAB/Simulink, Gazebo);
• Familiarity with robotic manipulators and manipulator kinematics/dynamics.
• Digital Twin Development and Simulation
• Experience in digital modelling of robotic systems and
  - physical integration;
• Knowledge of digital twin tools such as Unity, Unreal Engine, Siemens NX, and MATLAB Simscape;
• Ability to represent complex kinematic and dynamic models, including noise and sensor uncertainties.
• Applied Artificial Intelligence and Data Analysis
• Proficiency in multivariate analysis techniques (PCA, regression, clustering) and predictive risk models;
• Experience with artificial neural networks and fuzzy systems for KPI aggregation;
• Knowledge in data engineering, Python, and
  - time data analysis.

Generic Scientific Area: Electrical Engineering or Computer Engineering

Specific Scientific Area: Automation and Robotics

Requirements:

Master’s degree in Electrical Engineering or related fields; enrolled in a Ph
D program or a
- degree course.

The grant has a duration of 12 months, possibly renewable for identical periods on an exclusive basis, according to the Research Grant Regulations of the Polytechnic Institute of Coimbra – Order No. 5963/2020, of 01/06, with expected start after 16th January 2026.

Master’s degree in Electrical Engineering or related fields; enrolled in a Ph
D program or a
- degree course.

Experience in:

a) ROS, programming in C++ and Python;

b) Modelling and control.

Selection process

Selection Criteria

• Curriculum Evaluation (70%) and Interview (30%).
  • Curriculum Evaluation: Overall merit of CV (70%);
  • Publications in conference proceedings and journals (30%);
  • Interview Evaluation:
    • Knowledge and suitability for the role (50%);
    • Availability and schedule flexibility (20%);
    • Motivation and interest (30%)
#J-18808-Ljbffr