Antenna Range Evaluations of Low THz Imagers for Automotive Applications David R Vizard VivaTech Consulting SARL Nice, France drv@vivatech.biz M. Gashinova, E.G.Hoare, D.Jasteh, L. Daniel, M.Cherniakov EECE, University of Birmingham, UK m.s.gashinova@bham.ac.uk T-Y Tran, N. Clarke Jaguar LandRover, Research Department Coventry, UK Abstract—In this paper the requirements for low-THz automotive sensors are presented with the focus on THz imaging of the terrain in front of the vehicle. Initial imaging performance at 150 GHz radar is demonstrated and a planned extension using an advanced 300 GHz radar system is described. Antenna requirements for these systems are discussed. I. INTRODUCTION Autonomous navigation and ‘hands-off’ driving are anticipated modalities of future vehicles and mobile robots, where smart sensors and advanced processing technologies must establish vehicle “cognition” under all weather conditions. This must include all possible environments for safe driving and the protection of road users. To reinforce the capability of the autonomous platforms, new sensing systems are required to enable automatic optimisation of the vehicle during travel over any terrain. The most promising technology to enable these functions are electromagnetic remote sensing systems operating in the TeraHertz portion of the spectrum. The worldwide automotive market will need a new generation of sensor components, subsystems and systems, where application specific antennas and radar systems will play a key role. . II. STATE OF THE ART IN AUTOMOTIVE SENSING A. Low-THz automotive application The University of Birmingham, UK, Jaguar LandRover and VivaTech SARL, France, are undertaking collaborative research into low-THz radar system designs for automotive applications in the area of driver assistance. Providing driving assistance requires the implementation of fundamentally different tasks, which in the ideal case would be accomplished by a single sensor system. These tasks include terrain mapping, road surface identification, obstacle avoidance, pedestrian, child and animal recognition and tracking. Classification and identification of the object is the key feature of any modern sensor system which allows appropriate measures in specific situations, for which for the majority of applications rely on the extraction of imaging and specific physical features. A new generation of automotive sensors capable of operating in harsh open environments is required and low-THz frequencies (0.3-3 THz) are probably the best candidate to provide the required capabilities for automotive use. Low-THz sensing is expected to deliver information-rich imaging combining the advantages of electro-optical technologies in terms of resolution with the ability of radiowaves to propagate through adverse environments (e.g. spray, fog, dust, snow, etc.) [1,2]. In addition to conventional imaging, the radar capabilities can provide additional features, such as techniques to augment the image with parameters such as the range to the object, depth of shadow, resonances in the backscatter signal, and the ability to extract classification features that will imbue the image with texture, object curvature, and motion parameter estimation. Sensors for automotive applications are fundamentally relatively short range devices, with less than 200m range, and therefore atmospheric attenuation over a 200 m path is not a significant issue as attenuation in either rain, drizzle or fog for frequencies above 100 GHz up to 3 THz is relatively constant and does not exceed 10 dB per km [1]. Current automotive sensing systems include ultrasonic sensors, 24 and 77 GHz radar sensors, laser sensors, optical, and IR cameras. Each has fundamental drawbacks: electro- optical sensors fail to deliver a usable image under the most trying road conditions with poor visibility: fog, spray etc., ultrasonics provide short range sensing with the inherent limitation of contamination of the acoustic signal in presence of wind, exhaust, air brake noise etc., whilst traditional radar, robust to such conditions, demonstrates very limited imaging capabilities. The automotive applications of Low-THz must address the inherent problems of operational scenarios: (a) eliminating the operational difficulties of existing imaging technologies in harsh road environments, (b) producing high-resolution, information-rich road environment images from a single imaging technology (avoiding data fusion issues, such as ambiguity resolution and computational overheads arising from the use of multiple imaging technologies) and (c) achieving these in real time, unlike existing static THz imaging systems that require inordinately long integration (measurement) times in order to deliver high-resolution images [3,4].