From 2018 to 2022, a total of 375 million radars will be installed on new cars. What problems can arise with these systems?
Now that the Regional New Car Assessment Program (NCAP) requires Adaptive Cruise Control (ACC) and Emergency Braking (AEB) to assign its five-star safety ratings, NXP Semiconductors is encouraging the automotive industry to work on radar systems.
Sensors integrated in the vehicle
To accelerate the integration of radars into modern ADAS, on Tuesday (October 2), NXP released a solution combining S32R processors, an RF transceiver and an antenna on a new reference platform. Developed in partnership with Colorado Engineering, this platform meets “the stringent requirements for industry functionality, performance and security,” NXP said.
The new system was designed to dispel the myth of the “intricacy” of radars, which usually requires large automotive OEMs to fine tune the antenna and analog designs. NXP hopes that its “non-standard” car radar system will be able to serve Chinese automakers who still need a few years to catch up with the car market in the rest of the world.
In a recent EE Times telephone interview, Kamal Khouri, vice president and general manager of ADAS at NXP, told us: “Radar has become the most preferred sensor” for ACC and AEB. “Cameras cannot measure speed, unlike radars,” he explained. “Thanks to the reflection of signals, radars can also see around corners. On the other hand, lidars that do not use moving parts are still very expensive. ”
However, it is well known that the traditional radar lacks permission, and also it cannot distinguish nearby objects. Moreover, radars are notorious for having false positives, and they do not process information fast enough to be useful on the highway.
Khoury made it clear that the NXP does not believe that radars will replace cameras. “The combination of cameras and image radars provides redundancy, which makes cars safer,” Khuri said.
Analysis of a new radar solution
So what does the new NXP solution entail?
The reference design, dubbed RDK-S32R274, combines the NXP S32R27 processor, the TEF810x CMOS transceiver, the FS8410 power management chip, and the radar software development kit. NXP has added expansion modules and antenna modules that can be optimized to create a customized development platform for specific client applications.
The radar solution is based on a scalable family of processors based on Power Architecture – S32R27 and S32R37, which Khuri described as “the first chips designed to process radar algorithms.”
Block Diagram S32R NXP
According to Roger Keene, manager of radar at ADAS for automotive microprocessors, IP processing in NXP radars is performed on their own processors, in addition to automotive-grade software for ACC and AEB. Board and antenna modules designed for the company’s radar solutions are “reliable as certified automotive systems.”
Using the NXP’s automotive radar SDK, developers who used to manually configure their own IP radar processors for specific hardware can now use the features of the NXP radar system, Keen explained.
The S32R27-based solution is designed for advanced applications such as ACC and AEB. S32R37, with less processing power than S32R27, is compatible with the source code and optimized for operations such as detection of blind spots.
The cost of the S32R27 version is $ 14-17 (the price is when you buy 1000 modules). The cost of the solution based on S32R37 is 10-12 dollars.
Integration of automobile radars is not only the responsibility of NXP. Ian Riches, Executive Director of Global Automotive Practice at Strategy Analytics, considers NXP and Infineon among the leaders in automotive radar.
Meanwhile, Texas Instruments, which recently entered the radar market, began to catch up with the market in 2017, introducing millimeter-wave radar chips built on its own standard CMOS RF technology. TI told us that its radar chips provide “a resolution accuracy of less than 5 cm, a detection range of up to hundreds of meters and a speed of up to 300 km / h.” An even more important factor that distinguishes TI is that their microcircuit combines wave devices using mmWave technology with a 76-81 GHz wave radar, a microcontroller (MCU), and digital signal processor (DSP) cores on a single chip.
TI chose this approach because a higher level of embeddedness can reduce footprint, power consumption and cost without sacrificing performance. Cédric Malaquin, market analyst for radio frequency devices and technology at Yole Développement, told us that while NXP took the first step by developing its RF transceiver based on RF-CMOS technology, TI went further by integrating DSP into your radar chip. Malakin claims that DSP integration allows TI to reduce its footprint by almost 60%. DSP is the key to the “signal processing chain for detecting and classifying objects.”
However, NXP defended the company’s two-chip solution (radar chip + microprocessor), emphasizing that this approach offers customers much more scalability and flexibility for radar integration.
NXP Radar Solution: Antenna Side
NXP Keen said the following: “Consider operating at 43 ° C in Arizona.” He also said that the location of transceiver chips away from the microprocessor, for example, makes it easier to control thermal conditions when radars are installed in bumpers.
NXP Radar Solution: CPU Side
Keen also added that the NXP approach – the use of processors specifically designed for IP radar processing increased the performance per watt for radar solutions. Under pressure from the benchmark used to analyze performance per watt, NXP said it was gathering information “from open data” and “confidential customer meetings.” But Keen added, “Although we achieved the best performance per watt we have ever seen, we have consolidated the industry’s broader testing requirements with third-party companies.”
In response to a request to compare TI chips with NXP solutions, Riches of Strategy Analytics noted that “in essence, the TI approach could potentially offer lower cost, but at the same time, slightly less flexibility.”
Radar suppliers and market research firms are optimistic about the growing demand for car radars.
Different applications of radars require the creation of many different radar modules. NXP told us that “typically two radar modules are used in the two rear corners of a vehicle to detect blind spots. In more advanced tasks (such as detecting cross-movement), two more radar modules are required for the front corners of the car. ”
NXP claims that when using long-range radars, one module is usually installed somewhere in the front bumper.
Strategy Analytics predicts that a total of 375 million radars will be installed on new passenger cars from 2018 to 2022. Riches believes 107 million radars will be installed in 2022.
NXP Radar Market Ratings by Application
According to NXP estimates, in 2022, 109.2 million units of radars will be delivered – from corner to high-tech corner and long-range / medium-range models, including front / rear radars, which led to the introduction of radars in 50% of all new cars.
Radars Building Images
The newest trend among new radar solutions is how the most efficient radar systems can generate a high-resolution “image” by which you can both determine the location and identify / classify objects in the field of view. according to Riches of Strategy Analytics, “modern radars used in vehicles do not have the required resolution, which will allow you to form the correct image with a sufficient viewing width.”
This goal cannot be achieved only with the help of radar chips. Riches explained: “The antenna design is very important, and this is one of the reasons we saw startups like Metawave receive funding from companies like Infineon, Denso, Toyota AI Ventures, Hyundai Motor Company and Asahi Glass (among others).”
The advantages of radar technology are well known, especially their ability to work in any weather conditions. Automotive experts believe that radars can work with computer vision sensors and form a bunch to detect critical situations in highly automated vehicles.
Riches from Strategy Analytics explained:
In fact, they operate at very different wavelengths. The cameras (obviously) use visible light, and therefore they work worst in the dark, in very high contrast conditions (for example, when exiting the tunnel) or in heavy rain / snow. Lidars emit light outside the normal visible spectrum, but have the greatest problems in bright sunlight, which gives the system a lower signal to noise ratio. High-resolution lidar technology is now also expensive and less mature in the automotive industry than cameras or radars.
In turn, he noted that the radars are “immune to lighting conditions, while they have good penetration during rain or snow.”
However, radar is not the ultimate solution. The main disadvantage of the radar today is its low resolution: “he knows how to say that the object is present, but he will not be able to recognize this object,” Riches said.
Simply put, radar technology may not be appropriate in order to “make an informed decision about whether to continue driving (for example, an elevated street sign has been detected) or to perform emergency braking (a fire engine is parked in front of the traffic lane).”
All this explains why modern car radars sometimes discard and ignore stationary objects. “The radar cannot determine if an object is something you don’t want to crash into,” complained Riches.
In fact, the operating manuals are full of warnings for drivers whose vehicles are equipped with radar. Riches gave some examples.
The following text is taken from the manual for Skoda Superb (which uses radar-based ACCs):
“ACC does not respond when approaching fixed obstacles, such as traffic jams, damaged vehicles or standing at traffic lights.” (P. 236)
Volvo XC90 user manual contains similar warnings:
“The Distance Alert is active at speeds above 30 km / h (20 mph) and only responds to cars moving in front in the same direction as your car. Distance information is not provided for oncoming, slow-moving or stationary vehicles. ” (Page 289)
“The Pilot Assist does not brake in front of people, animals, objects, small vehicles (such as bicycles and motorbikes), low trailers, or oncoming, slow or stationary vehicles.” (P. 310)
Riches concludes: “You will find similar text in many other operating manuals from many other brands. The purpose of image building radars is to fix this problem. ”
We are perhaps the most powerful competence center in Russia for the development of automotive electronics in Russia. Now we are actively growing and we have opened many vacancies (about 30, including in the regions), such as a software engineer, design engineer, lead development engineer (DSP programmer), etc.
We have many interesting challenges from automakers and concerns driving the industry. If you want to grow as a specialist and learn from the best, we will be glad to see you in our team. We are also ready to share expertise, the most important thing that happens in automotive. Ask us any questions, we will answer, we will discuss.
Read more useful articles:
- Free Online Courses in Automotive, Aerospace, Robotics, and Engineering (50+)
- [Прогноз] Transport of the future (short-term, medium-term, long-term horizons)
- The best materials for hacking cars with DEF CON 2019-2020
- [Прогноз] Motornet – a data exchange network for robotic vehicles
- Companies Spend $ 16 Billion On Unmanned Vehicles To Capture 8 Trillion Market
- Cameras or lasers
- Autonomous cars on open source
- McKinsey: Rethinking Software and Electronics Architecture in Automotive
- Another OS war is already under the hood of cars
- Program code in the car
- In a modern car, there are more lines of code than …