Radar, a promising future

     Radar applications are expanding beyond traditional uses, driven by new regulations and advances in radar technology. This trend is occurring not only in established industries such as automotive, industrial, and defense, but also in new areas such as consumer electronics and healthcare. Improvements in accuracy, size, cost, and power consumption make radar a good fit for these new markets, opening the door to huge opportunities.
     Overall, the radar module market is expected to reach nearly $36.5 billion by 2029, representing a 5% compound annual growth rate from 2023 revenues of $26.7 billion.

Some Trends in Radar
   The shift from GaAs to SiGe and now CMOS radar modules is groundbreaking in the rapidly evolving world of automotive technology. This advancement has not only slashed costs, but also foreshadows that the price of a standard 77GHz radar could plummet to $30 by 2030. There is a concerted effort to reduce the average selling price of cutting-edge 4D imaging radars to accommodate OEM budget constraints.
Despite advances in radar technology, current capabilities still fall short of the demands of fully autonomous driving. Module manufacturers are actively exploring various architectural approaches to bridge this gap, aiming to expand the field of view and improve angular resolution. The experimentation involves adding RF channels, increasing computing power, or mixing both strategies for the best results. It’s a vibrant outlook that fosters innovation in pursuit of the ultimate goal of autonomous driving.
   In addition, OEMs are leading a major shift toward vehicle centralization, which is expected to be fully realized between 2030 and 2035. This shift foreshadows a future where more affordable and compact radars with enhanced computing power and superior system performance will become standard for a truly autonomous driving experience.

   As mentioned in last year’s report, OEMs are increasing the rate of radar in cars. Several developments are happening or expected to happen. This includes switching from 24 GHz to 77 GHz, moving away from traditional radars without elevation capabilities, and moving from a limited list of tracked objects to 4D radars as baseline and imaging radars in advanced cases. There is also a trend towards centralizing radar computation and moving from planar PCB antennas to 3D waveguides.
   In addition to external radar sensors for driver assistance, car interior monitoring is also gaining attention. The first implementation is driver monitoring systems to ensure that the driver is focused on the road ahead. Starting with child presence detection, car occupant monitoring systems are a natural extension of passenger safety. Although it can also be used to improve the user experience. Next comes object monitoring, such as the position of seats or headrests.
CPD systems have become a requirement in several markets (ASEAN NCAP, Euro NCAP) since 2022/2023, but the system itself is not regulated. In most cases, indirect methods (such as door opening cycle tracking and driver alerts) are used. However, starting in 2025, Euro NCAP will make direct sensing mandatory, which could change the market dynamics for in-car monitoring.
   Radar is particularly well suited for this task, as it can detect children in flip-up baby seats. It can also be used for vital signs monitoring.
   Meanwhile, automotive radar sensors are beginning a paradigm shift, as Yole said in a report last year. Their task has evolved from tracking a limited list of moving objects to generating perception maps.
The first breakthrough improvement is the use of radar modules (the fourth “D” of radar) for altitude measurement. This is key to deciding whether to drive over road debris and under bridges, and is the focus of the fifth generation of radars from leading manufacturers.
   But the most important breakthrough required is an order of magnitude increase in the angular resolution required for correct target separation. The first so-called imaging radars achieved 1° angular resolution by extending the MIMO concept. The principle is to increase the number of transmitting and receiving antennas to obtain a larger virtual antenna array aperture. However, there are some physical limitations to antenna scaling, first of all the size of the array. Another limiting factor is the computing power and memory resources required for such an array. The solution may be computing centralization. With a centralized architecture, the computational part of the radar may be removed and moved to a regional ECU. The radar will become cheaper and smaller, and its computing power will be enhanced, thereby improving its performance.
   Again. Vehicle centralization is a new trend for OEMs and is expected to become a reality around 2030-2035. The performance of the RF sensor itself has improved significantly. Key quality factors have improved and temperature stability is better. At the same time, the integration level has been further enhanced due to the development towards mature CMOS technology.

Increasing competition forces chipmakers to reassess their strategies
​  data shows that in 2022, six major players still dominate the radar market: Continental, Aptiv, Bosch, Hella (now Forvia), Denso and Veoneer (later Magna). However, these companies face challenges in customer willingness to invest in software-defined vehicles. They also face growing competition from China, such as SAIC's Hasco and BYD's FinDreams Technology. Many have received funding since 2015.

   At the semiconductor level, the market in 2022 is dominated by the duopoly of NXP and Infineon. TI is the largest competitor; Renesas is just now positioning itself with a full portfolio, while Arbe has gained market acceptance and started generating revenues as its solutions are accepted by Chinese Tier 1s and a prominent European Tier 1 (Veoneer). Another important player ready to seize the automotive radar market opportunity is MobilEye. The company plans to offer commercial radar services on its next-generation “Chauffeur” by 2025.
  Most semiconductor radar players are positioned for edge processing, which is the main market today and will likely be the main market for the next five years. However, as mentioned above, the shift to centralized radar computing has already begun.
  The radar module market is highly competitive, with numerous companies vying for market share. Competition is intense in all segments, from the crowded automotive radar to the specialized field of industrial applications. As a result, module suppliers are constantly pushing prices down to remain competitive. The recent momentum of collaboration between OEMs, Tier 1 suppliers, chipmakers, and antenna manufacturers reveals the strategic stance taken by each player.
  While the module supplier landscape is fragmented, the radar equipment market is highly competitive, with only a few players in the lead. These incumbents, who typically supply across industries, are now facing a new wave of competition from completely different players, from consumer electronics companies to OEMs and Tier 1 suppliers looking to climb the value chain by developing their own chips.
  Recent technological advancements are expected to leave their mark on the market this year, heralding a major shift in the supply chain landscape. CMOS-based radars integrate transceiver and processing functions into a single radar chip, promising cost savings, space efficiency, and supply chain simplification for OEMs. In addition, automotive companies are preparing for a transformational phase in electronics/electronic architectures that will fundamentally redefine the roles of industry players.