RAPTAR: Radar Radiation Pattern Acquisition through Automated Collaborative Robotics

TL;DR

RAPTAR is an autonomous robotic system that enables precise 3D radiation pattern measurements of radar modules in diverse environments without specialized facilities, improving accuracy and ease of testing.

Contribution

This work introduces RAPTAR, a portable robotic system that automates radiation pattern acquisition, overcoming limitations of traditional probe-station methods and enabling measurements in real-world settings.

Findings

Achieves angular resolution up to 2.5 degrees.

Mean absolute error of less than 2 dB compared to simulations.

Demonstrates 36.5% lower error than baseline methods.

Abstract

Accurate characterization of modern on-chip antennas remains challenging, as current probe-station techniques offer limited angular coverage, rely on bespoke hardware, and require frequent manual alignment. This research introduces RAPTAR (Radiation Pattern Acquisition through Robotic Automation), a portable, state-of-the-art, and autonomous system based on collaborative robotics. RAPTAR enables 3D radiation-pattern measurement of integrated radar modules without dedicated anechoic facilities. The system is designed to address the challenges of testing radar modules mounted in diverse real-world configurations, including vehicles, UAVs, AR/VR headsets, and biomedical devices, where traditional measurement setups are impractical. A 7-degree-of-freedom Franka cobot holds the receiver probe and performs collision-free manipulation across a hemispherical spatial domain, guided by real-time motion planning and calibration accuracy with RMS error below 0.9 mm. The system achieves an angular resolution upto 2.5 degree and integrates seamlessly with RF instrumentation for near- and far-field power measurements. Experimental scans of a 60 GHz radar module show a mean absolute error of less than 2 dB compared to full-wave electromagnetic simulations ground truth. Benchmarking against baseline method demonstrates 36.5% lower mean absolute error, highlighting RAPTAR accuracy and repeatability.