Thesis Code: 24020
Research Area: Connected Systems and Cybersecurity

Motivation
In recent years, the autonomous operation of mobile robots has significantly progressed. These robots are now used in various domains, from industrial automation to exploring dangerous environments. One major challenge for these systems is to autonomously operate in GNSS-denied environments. Therefore, having robust and reliable localization methods, alternative to GNSS, is important.
Recently, Visual Odometry (VO) is attracting increasing attention because of its lightweight, accuracy, and reliability, which has been successfully applied to many real-time robotic systems. However, VO methods suffer from localization drift over long trajectories due to the inherent limitations of visual-based localization methods. To address this issue, recent research has explored the integration of VO with Ultra-WideBand (UWB) technology and Inertial Movement Unit (IMU) sensors.

Objectives
The goal of this thesis is to improve the localization accuracy of VO solutions by integrating data from both UWB positioning technology and IMU sensor. Several integration approaches, of different complexity, can be found in literature. In this thesis a Factor Graph (FG) framework will be adopted to perform such data fusion. FG describes positioning problems in terms of optimization problems, allowing the solution to be obtained over multiple iterations, differently from other traditional navigation filters such as Extended Kalman Filter (EKF).

The thesis activity will be focused on the design and development in Matlab environment of a positioning algorithm, based on the FG framework, combining VO, IMU and UWB measurements. In particular, UWB ranging measurements will be performed by a mobile robot with respect to fixed UWB devices, called UWB anchors, whose positions are well known.
Initially, the designed FG algorithm will be tested through computer simulations. Subsequently, real measurements will be collected by a rover platform (equipped with VO, UWB and IMU sensors) moving in the Robotics Laboratory made available by LINKS. The collected measurements will then be used to test and optimize the algorithm off-line.

Requirements
– Strong knowledge of MATLAB
– Good knowledge in software development (C/C++)
– Good knowledge in mathematical derivation
– Proactive mindset, problem-solving oriented
– Data processing skills

Duration: 6-8 months.

Contacts: please send a resume with attached the list of exams passed during the Bachelor of Science and Master of Sciences to francesco.sottile@linksfoundation.com specifying the thesis title.