###Through-the-Wall Radar Imaging and Mapping using Robots
Project ID: 2228bd1068 (You will need this ID for your application)
Research Theme: Artificial Intelligence and Robotics
UCL Lead department: Security and Crime Science
Lead Supervisor: Kevin Chetty
Project Summary:
State-of-the-art approaches for mapping the layouts of building interiors are based on optical technologies such as cameras and LiDAR. These systems must be physically located within the building with unobscured line-of-sight to key physical structures such as walls, ceilings, and stairwells. Although these technologies are finding many real-world applications, they are ineffective in life-critical scenarios where the generation of floorplans reflecting the existing state of the building structure, as well as the locations of people within them, are needed to provide first responders and/or law enforcement with real-time situational awareness. Examples include the collapse of buildings resulting from earthquakes or military attacks, hostage situations and terrorist events. Moreover, in these cases the mapping must take place at safe distances away from the building to aid decision making and tactical planning.
To address this technological gap, the PhD aims to develop a through-the-wall radar-based imaging robot for mapping the internal layouts of buildings and locating people within them. Its operational capability will also be assessed in field trails alongside the Metropolitan Police. Radar offers significant advantages over current solutions for indoor mapping owing to its ability to operate in low light, smoky and dusty conditions – which severely inhibit optical sensors. Moreover, they can “see” through solid walls providing opportunities to map interiors from stand-off (safe) distances.
A PhD student with a background in radio-frequency sensing, image and signal processing, and machine learning will explore methods for transitioning SLAM (Simultaneous and Localization and Mapping) techniques – predominantly used in machine vision - to radar. The proposed research therefore provides a fantastic opportunity to provide a step-change in radar sensing capabilities and spawn a range of new applications. To date, there are no comprehensive systems that can remotely map the inside of a building, whilst detecting and tracking people located within it.