Seeing the unseen: The power of 4D imaging radar

Insights
Technology Trends

Q&A with Infineon Technologies AG.

Four-dimensional imaging radar is an emerging key technology for autonomous driving, providing real time and accurate insights into the surrounding environment. It complements other sensors such as cameras and is used in conjunction with these in the automotive industry. The 4D imaging radar uses multiple antenna arrays to gather extensive data and generate precise 3D models. It operates at frequencies such as 77 or 79 GHz, enabling split-second reaction times that are required for autonomous vehicles. The future of imaging radar involves improving sensor fusion algorithms and developing wider aperture radar systems to achieve higher levels of autonomy in self-driving vehicles. To learn more, we spoke to executives of Infineon Technologies AG.

Key takeaways

  • The 4D imaging radar is a significant advancement in radar technology that enhances automated and assisted driving systems by accurately measuring elevation.
  • Various use cases necessitate varying levels of performance from 4D radars, including the detection of overhead bridges and powered two-wheelers. For early detection of road debris, higher performance radars are required.
  • By 2030, vehicles may require up to seven times 4D radars for Level 3 and Level 4 autonomous systems, working in conjunction with cameras and Lidar modules.
  • The 4D imaging radars and Lidar will coexist and complement each other in autonomous driving. Radars are reliable in challenging weather conditions, while Lidars offer better object mapping capabilities.  
  • Reliability is a key requirement for 4D or imaging radars, and Infineon prioritizes product quality, functional safety and expertise in radio frequency (RF) and signal processing.

The following is an edited transcript of the conversation.

S&P Global Mobility: In your own words, can you describe the significance of 4D or imaging radar in the evolution of radar technology?

Infineon Technologies: 4D imaging radar represents a significant advancement in radar technology, offering increased road user separability by a factor of [sixteen times], improved resolution by a factor of [five times], and the addition of elevation measurement capabilities. As a result, it plays a pivotal role in enabling the next stage of automated and assisted driving by effectively addressing critical use cases.

The need for 4D radar arises from the necessity to measure elevation accurately. Currently, conventional radar modules primarily utilize all channels in the azimuth, resulting in improved separability in the horizontal axes. However, the elevation values provided are only estimations, which adequately address present use cases but may not be sufficient for future requirements.

What new use case brings the first 4D or imaging radar to market?

Different use cases require varying levels of performance. Two highly debated applications for 4D radars in the market are the reliable and secure detection of overhead bridges and the detection of powered two-wheelers.

More advanced imaging radars will be necessary to detect road debris early on and facilitate trajectory planning.

To ensure the robust and safe deployment of the mentioned use cases, reliable radar modules with sufficient performance are required, starting from Level 2+ and its associated features such as automated lane change.

In what ways does imaging radar improve automated driving or enable higher levels of automation, such as in navigating tricky situations, managing false positives, or filling a performance gap in other common sensors [such as] cameras, traditional radar or Lidar?

[Four-dimensional] imaging radars offer enhanced separability, allowing for better differentiation of road users, such as an adult next to a truck or a bike next to a truck. This capability is particularly crucial in urban scenarios with diverse road users or during traffic accident situations. Moreover, a higher channel count in 4D imaging radars helps reduce sidelobe effects, resulting in a more reliable system with a lower false-alarm rate. Finally, radar is an optimal sensor for detecting distance and speed information. With the inclusion of Lidar-like angular information and improved road user classification, 4D imaging radar further enhances its capabilities.

Considering advances in automated driving and autonomy levels, how many 4D or imaging radars do you believe will be required in a vehicle in 2030?

For Level 3 and Level 4 systems, we expect vehicles to require up to [seven times] 4D or imaging radars with different levels of performance.

This implies that there will be a need for high-performance radar modules with a narrow field of view (FOV) and long range for forward and rearward detection. Additionally, short-range sensing with a wide FOV will be necessary to enable precise object detection in urban scenarios. These radar modules will work in conjunction with camera and Lidar modules in a sensor fusion mode to enhance overall system capabilities.

How can interior radar be utilized for occupant detection and health monitoring within vehicles?

In this case, we can leverage the inherent benefits of radar sensor technology. Interior radar uses range processing schemes to detect body volume and can also detect micromovements of the chest (such as breathing) and movements of extremities through Doppler shift processing.

These capabilities are achieved with a compact form factor and a user-friendly approach, allowing for rapid deployment of the features.

Can 4D or imaging radars potentially replace or augment Lidar sensors? Or are imaging radar and Lidar more likely to coexist in vehicles?

Currently, there is a specific requirement for Lidar technology in Level 4 and Level 5 autonomous driving. However, all other use cases and functions in Level 3 and below can be effectively addressed through the fusion of camera and radar technologies.

Based on this, we expect 4D imaging radars and Lidar to coexist and complement each other. [Four-dimensional] imaging radars are highly reliable in various weather conditions, from fog to rain to snow, where Lidars are challenged.

On the other hand, Lidars currently offer superior object/environment mapping compared to radars. Nevertheless, as imaging radars continue to improve in performance, offering more detailed detection capabilities, they will also be capable of mapping and classifying objects in the future.

How does imaging radar compare to traditional radar and Lidar sensors in terms of cost?

Imaging radars provide a clear cost advantage compared to Lidar. The cost of imaging radars depends on the configuration and can start with a [60%] cost increase compared to traditional radars. In contrast, the current generation of Lidar sensors relies on complex opto-mechanical-electrical systems that are not easily scalable in terms of cost. However, we have observed a strong cost reduction in radars over the past generations, despite the continuous performance improvement.

Please share any insight or commentary on 4D or imaging radar not otherwise addressed above.

We talked about the performance and use cases, but one topic that we need to really highlight is that one of the most important requirements for 4D or imaging radars is reliability. At Infineon, we believe that increasing performance has to be hand-on-hand with reliable concepts and sensors. We reflect our mindset in the definition of our products where product quality and functional safety are one of the key pillars together with excellent RF and signal processing know-how.

Imaging and 4D radar sensors are the latest innovation in the storied automotive radar sector, and many automakers are introducing them today and in their next-generation platforms in just a few short years. S&P Global Mobility estimates that production of imaging and 4D radar sensors will grow at a compound annual growth rate of nearly 10% through the end of the decade. To learn more, please visit Component Forecast Analytics or contact your account manager.

preload preload preload preload preload preload