Automated Guided Vehicles (AGV) are gaining in popularity because they can carry out many tasks, bringing efficiency and sometimes safety to a wide range of industrial, domestic and surveillance/security applications. AGVs typically embark a suite of sensors, each with distinctive advantages. Among these sensors, radars provide unique features of robustness in presence of dust, rain and extreme lighting conditions. Radars excel at measuring range and instantaneous velocity but are usually weaker to measure angular information with good resolution.
Imec envisions AGVs fitted with forward-looking, side-looking and backward-looking radars. Each of these radars must have high resolution in all dimensions (range, azimuth, elevation and velocity) so that the combined mapping information provide 360° coverage. To reach this goal, special techniques such as Synthetic Aperture Radar (SAR) and Doppler Beam Sharpening (DBS) need to be developed. This is illustrated in the figure below.
The aim of this thesis is:
- To develop the algorithms for DBS and SAR on an existing radar prototype. This prototype is a uniform linear array conceived for conventional beamforming. The concept is to use the linear array for the elevation resolution and the SAR or DBS for the azimuth resolution.
- To develop algorithms exploiting the images from the enhanced AGV radars to build an occupancy grid and to perform Simultaneous Location and Mapping (SLAM). SLAM is a technique whereby a sensor creates a mapping of the environment and locates itself in this map. Fusion of the information from the different radars must also be performed to produce the 360° coverage.
This work will be performed within a team working on all aspects of the radar, from RF to signal processing and machine learning. The work will include literature study, algorithm development and experimentation with a mm-wave radar.
We accept two students for this MS thesis, one working on the SAR/DBS algorithms and the other on the occupancy grid and SLAM algorithms.
The successful candidate(s) must show a good understanding and deep interest in signal processing and computer science. Proficiency with Matlab, Python and C/C++ is a must. Some knowledge of radar concepts is a plus.
Type of project: Combination of internship and thesis
Duration: 6 months
Preceded by optional summer internship (1 to 3 months) - (the summer internship alone is not possible)