In its latest report, "Radar Technologies for Automotive 2018," Yole Développement (Yole) investigated radar’s various evolution trends as well as its ecosystem and supply chain. Yole positions radar as a key technology for automotive sensing, with an increasing penetration rate.

The use of electronic components in automotive is exploding, as OEMs and Tier-1s focus heavily on the ADAS (Advanced Driver Assistance System) to deliver safer cars and reduce road fatalities. Safety bodies like Euro NCAP (Europe New Car Assessment Program) are pushing this way too, encouraging OEMs with high ratings for cars equipped with advanced safety functions such as AEB (automated emergency braking). Applications are still evolving with requirements for highways and cities, road intersection scenarios, and vulnerable road-user detection specific to pedestrians and cyclists. Today’s most advanced cars, which currently assist the driver as defined by automation levels 2 and 3, will progressively upgrade to levels 4 and 5, leading to more accurate sensor integration. Numerous sensors have been developed to serve as a car’s “eyes”, and support automation. The camera sensor is a natural technology of choice for this task, with its object recognition capability. However, it has range limitations (100 m best-case) and struggles to work in adverse weather conditions. Thus, other sensing technologies are required to enable further car automation. Radar technology fills the gap, since it is able to detect objects up to 250 m in front of the car, even in fog and poor visibility. Radar also has an impressive technology roadmap that allows for huge range and angular resolution improvement, as well as device miniaturization and cost reduction. It is also well suited for accurate velocity extraction.

Yole believes radar technology will achieve an outstanding penetration rate in car sensors complementing camera devices. Despite small growth (~3 percent) in global car sales until 2022, Yole Développement expects an average growth rate of 23 percent for radar module sales, and an average growth rate of 22.9 percent for radar chip sales over the next five years--with autonomous driving being the next long-term driver for radar technology growth.

Automotive radar operates at 24 GHz in the unlicensed ISM (industrial-scientific-medical) band for short-range (up to 30 m) applications (blind-spot detection, lane-change assist) and at 77 GHz in the W-Band for long-range (up to 250 m) applications (adaptive cruise control, automated emergency braking). However, this heterogenic approach might generate interference issues with an increasing number of radar-equipped cars, and so a more unified platform called 79 GHz has been proposed, based on the 77 GHz frequency with its 5 GHz of available bandwidth from 76 to 81 GHz. 79 GHz offers other advantages too: it improves radar resolution to enable better target separation, while reducing antenna and high-frequency circuit size.

Based on this new platform, radar architectures will reach a new level of complexity, requiring innovation in antenna design, complex modulation techniques and target-resolution algorithms. Multi-beam, multi-range approaches lead to more complex antenna arrays that multiply transmit-and-receive paths while adding 3D detection capability. Imaging capability, which is the only thing radar lacks, is envisioned too, and radar-based developments for object classification are under way.

To support these stringent requirements, a new chip generation is needed with increasing channel numbers and integration of the analog-to-digital converter, as well as digital signal processing, together with the radar front-end on a single chip. A battle is underway between the well-established SiGe technology and the more recent RFCMOS platform, which is quickly becoming a reality thanks to players like Texas Instruments--which has spent the last decade developing RFCMOS technology.

Innovative startups like Metawave and Uhnder, which propose disruptive technologies for very high-resolution electronic steerable antennas and imaging radar, are competing head-to-head with well-established module makers like Continental and Bosch. Regarding automotive 77 GHz radar chips based on a 130 nm SiGe platform, NXP and Infineon are the top suppliers, with other big semiconductor companies like Texas Instruments and ADI offering products based on advanced CMOS nodes (down to 28 nm).

Foundries are also positioning themselves in this ecosystem. For example, GLOBALFOUNDRIES and its 22FDX platform, TOWERJAZZ and its 180 nm SiGe platform and UMS too. It is exciting to see such a wide diversity of technology offerings, a clear confirmation of the automotive radar market’s traction. However, penetrating the automotive market with new technologies is no easy task. On the contrary, entering and maintaining a position in the automotive supply chain is a long, trust-based process.

We are certainly entering a new “radar age,” with many developments, disruptive technologies and new entrants positioning this technology as the primary sensor--along with imaging (cameras) for ADAS and autonomous vehicles.