Autonomous control of a vehicle makes driving safer and comfortable. There are several critical technologies behind safe and efficient autonomous-vehicle operation—AI, safety and security, cameras, network infrastructure, and the sensor technologies radar and lidar, or laser-light radar. All these technologies must integrate seamlessly to help ensure safe and successful autonomous-vehicle operations.
In the automotive space, the primary radar applications can be broadly grouped into corner radars and front radars. Corner radars (at both the rear and front corners of the car) are typically short-range sensors that handle the requirements of blind-spot detection (BSD), lane-change assist (LCA), and front/rear cross-traffic alert (F/RCTA), while front radars are typically mid- and long-range radars responsible for autonomous emergency braking (AEB) and adaptive cruise control (ACC). In the industrial space, the applications for radar include fluid and solid level sensing, traffic monitoring, robotics and more.
The two frequencies commonly used in these radar applications are 24GHz and 77GHz. However, there is a shift in the industry toward the 77GHz frequency band due to emerging regulatory requirements, as well as the larger bandwidth availability, smaller sensor size and performance advantages.
Magna Reveals Autonomy-Enabling Radar
Magna has introduced a military-derived high-resolution radar system, the Icon Radar that enables precise image detection at more than 1,000 feet (about 300 m), continuously scanning to determine distance, height, depth and speed, says Swamy Kotagiri, chief technology officer-Magna.
Kotagiri says Icon Radar tracks nearly 100 times more objects than current radar systems, can distinguish between static and moving objects and can differentiate objects such as vehicles, bicycles, pedestrians and animals. The next-generation radar isn’t hampered by weather conditions or other interference.
Magna says the system scans 50 times faster than a human eye can blink, giving the vehicle constant information on complex surroundings, enabling instant decision making. It also can differentiate smaller, closer objects even when larger, more distant objects might reflect a stronger signal. Magna and technology startup Uhnder are engineering and validating the system with a plan to have it market-ready by 2019 and installed in production vehicles in 2020.
Mobileye has developed the sensors and software that allow a car to know where it is in relation to its surroundings. Mobileye-powered autonomous Ford Fusion car is equipped with 12 cameras, radar and scanners that give it a 180-degree view from a distance of up to 300 meters. The technology – dubbed REM (for “road experience management”) – uses Mobileye sensors to draw high-definition maps of road conditions in near real time, crucial for both fully autonomous driving and the advanced safety systems of today’s cars.
Radar Research could offer breakthrough for autonomous technology
A novel approach to radar technology developed by Tel Aviv University researchers has found that, contrary to popular belief, radar accuracy is not necessarily dependent on the range of frequencies or bandwidth in use. Breaking with long-held principles guiding the development of radar technologies for years, notably linking greater frequency range to increased accuracy, the scientists’ unorthodox method – inspired by Optical Coherence Tomography – enables precise and high-resolution mapping of the surrounding environment with little to no bandwidth. The new concept could provide solutions in fields including the autonomous car industry, optical imaging and astronomy where accuracy is required but bandwidth is limited.
The study was led and conducted jointly by Prof. Pavel Ginzburg, Vitali Kozlov, Rony Komissarov and Dmitry Filonov, all of Tel Aviv University’s School of Electrical Engineering. The findings have been published in the scientific journal Nature Communications.
“Here we demonstrate a different type of ranging system, which possesses superior range resolution that is almost completely free of bandwidth limitations,” said Prof. Ginzburg.
“This new technology has numerous applications, especially with respect to the automotive industry. It is worth noting that existing facilities support our new approach, which could be launched almost immediately to existing platforms to outperform outdated solutions.” By exploiting the coherence property of electromagnetic waves, the researchers said, low bandwidth radars can achieve similar performance to high bandwidth radars, but at a lower cost and without broadband signals.
Their unconventional “partially coherent” radar has experimentally demonstrated an improvement over an order of magnitude in resolving targets, compared to standard coherent radars with the same bandwidth.
While today not many cars on the road use radars and, as a result, there is almost no competition for allocated frequencies, what will happen in the future, when every car will be equipped with a radar and everyone will demand the entire bandwidth?” said Kozlov.
“We will find ourselves in a sort of ‘radio traffic’ jam. Here, achieving better performances with smaller bandwidth offers a necessary solution, allowing us to share the bandwidth without any clash.”
Research was carried out at the Tel Aviv University Radio Physics Laboratory’s anechoic chamber, and supported by a European Research Council grant and the KAMIN research incentive program. “Our demonstration is just the first step in a series of new approaches to radio-frequency detections, exploring the impact of low bandwidth radars on traditional fields,” said Prof. Ginzburg. “We intend to apply this technology to areas previously unexplored.”
Short-range Radar Reference design
TI offers a reference design, the TIDEP-0092, which can be used for short-range radar (SRR) automotive applications. The design features the AWR1642, an integrated single-chip FMCW radar sensor capable of operation in the 76-81 GHz band. The design is intended for short-range applications (that is to detect as many as 200 objects up to a distance of 80 m (260 feet) and track as many as 24 of them traveling as fast as 90 kph, which is approximately 55 mph). In the short range application, the sensor can simultaneously track objects at 80 m while generating a rich point cloud of objects at 20 m, so that approaching vehicles and closer small objects can be detected at the same time. A range of more than 80 m can be achieved with the design of an antenna with higher gain than the one included in the AWR1642.
Traffic Monitoring Object Detection and Tracking Reference Design Using mmWave Radar Sensor
The TIDEP-0090 reference design from TI uses single-chip millimeter-wave (mmWave) technology used for robust, long-range sensing in traffic monitoring and other applications. The design uses the IWR1642BOOST evaluation module (EVM) and integrates a complete radar processing chain onto the IWR1642 device. This design is also useful for building other applications, such as object detection for robots.
- Demonstration of environmentally robust object detection, clustering, and tracking using IWR1642
- IWR1642 mmWave sensor pinpoints location of objects over a range of 70m for multi-lane monitoring and 195m for single-lane monitoring
- Measurement bandwidth of 76 GHz to 77 GHz
- IWR1642 with an integrated digital signal processor (DSP) to cluster objects and track their range and velocity over time
- Based on proven EVM hardware designs enabling quick time to market and out of the box