Skip to main content Skip to navigation

Sepeedeh Shahbeigi Roudposhti

Sepeedeh

PhD in Engineering

 Sepeedeh.Shahbeigi-Roudposhti@warwick.ac.uk

Supervisors: Professor Mehrdad Dianati and Professor Paul Jennings

Research Interests:

Relative localisation, SLAM, Visual odometry

Biography

Sepeedeh is a PhD candidate Supervised by Professor Mehrdad Dianati and Professor Paul Jennings.

Sepeedeh received her B.Sc. in Aerospace Engineering from Sharif University of Technology, Iran, and M.Sc. in Mechanical Engineering from Koc University, Turkey. She is currently pursuing her Ph.D. degree with the Warwick Manufacturing Group.

Her research is focused on integrity monitoring of perception-based localisation solutions in automated driving systems. Sepeedeh’s research aims to address the safety issues faced in level 3 and above autonomy systems (SAE standard) by developing a method to monitor the integrity of local positioning solutions. Sepeedeh is a member of Royal Institute of Navigation – Younger Member Group.

My Research

Integrity is a performance metric of the localisation module and directly relates to the safety of autonomous vehicles. Systems need to be designed to monitor the position integrity and whenever it is not guaranteed, an alarm should be raised to inform the user. The integrity monitoring methods designed for autonomous vehicles localisation are adapted from the ones developed for civil aviation, where Global navigation satellite systems (GNSS) is the primary means of localisation. These methods commonly consist of two elements: 1) detection and exclusion of faulty measurements and 2) estimation of an upper-bound on the position errors. Applying the airborne integrity monitoring methods to automated driving systems faces various challenges such as dealing with prevalent multipath and non-line-of-sight interferences in the reception of the GNSS signals due to the crowded structure of urban environment. In addition, GNSS signals has been shown to be insufficient to satisfy the integrity requirements for automated driving systems. These issues raised the need to incorporate other sensors on board the automated driving systems such as cameras, LiDARs and maps into integrity monitoring methods. However, this research area is very recent and on-going.

My research goals are:

  • Investigate applicability of RAIM or similar approaches to selected relative localisation techniques that rely on perception sensors.
  • Design a bespoke integrity monitoring framework for relative localisation technique that can be used in highway driving scenarios and performed on realistic simulations.
  • To carry out a field test to develop proof of concept for field trial and evaluation of the proposed techniques.

Current Research Focus:

  • Integrity monitoring of perception sensor data
  • Relative localisation
  • ORBSLAM visual localisation
  • Visual odometry

Hobbies:

  • Hiking
  • Camping

hike