Did you know that the single-chip radar system-on-chip (SoC) is becoming one of the most popular new sensors? Its widespread adoption in cars has greatly increased the sales volume, and in turn lowered the price. These sophisticated ICs have become essential to automakers, but they’re also attractive for other applications. And though automotive applications of these ICs will continue to dominate, designers are now exploring a range of new uses that can improve safety and convenience.
- TI’s smart sensors ideal for automated driving applications
- Robust traffic and intersection monitoring using millimeter wave sensors
- mmWave sensors make cities smarter
Single-chip radars—who would have thought? Yet they now come in a variety of forms from multiple manufacturers. The chips have been developed for the 24-, 76- to 81-, and 94-GHz bands in most countries. Though both continuous-wave (CW) and pulse types are available, most use the frequency-modulated continuous-wave (FMCW) scheme at 77 MHz. Some of these devices are made with silicon germanium (SiGe), but the latest versions consist of CMOS or BiCMOS. Complete modules for auto or other industrial uses are offered by some companies. Their unique capabilities make them an attractive alternative to other sensors for certain new applications.
Most engineers tend to mentally categorize these devices according to their original use—as a radar system on a chip. However, it’s better to view them as just another type of sensor. Thus, millimeter-wave (mmWave) radars are an unexpected choice when seeking a device for proximity detection of objects, motion sensing, or making physical measurements.
1. This FMCW chirp signal is commonly used in the 76- to 81-GHz band.
Remember, radar is used to measure distance, direction (angle), and speed. Examples include police speed radars and baseball speed guns. A transmitter (Tx) in the chip radiates a signal that’s then reflected from a remote object and returns back to a receiver at the source. The transmitted signal is a linearly increasing frequency change for a short duration, called a chirp (Fig. 1). The chirps are repeated in a desired pattern.
Figure 2 shows the radar transceiver. The frequency of the return signal is mixed at the receiver (Rx) with the transmitted frequency to produce a different intermediate frequency (IF). That IF is digitized and used to determine motion and speed. The processing circuits on the chip measure the time of transit and compute the distance given the known speed of radio waves. And thanks to highly directional antennas, location (azimuth) can be detected. FM radar also lets you measure motion and speed. On-chip processors take care of the calculations to give you very precise measurement data. What you now have is a very flexible and programmable sensor for many unique applications.
2. Shown is a simplified diagram of a single-chip radar transceiver.
Radar Sensor Applications
The single-chip radar’s greatest application to date is automotive safety. Radars are at the heart of the advanced driver-assistance systems (ADAS) now available in most cars. Adaptive cruise control, automatic braking, backup object detection, blind-spot detection, lane-change assist, and cross-traffic alert all take advantage of these radars. The goal is to reduce driver error and, therefore, decrease the number of crashes, injuries, and fatalities. So far, that goal is being achieved. In fact, these new subsystems are so effective that the government is mandating ADAS for all cars.
Furthermore, such radars are absolutely essential to the success of the driverless car. They complement ADAS video cameras, LiDAR, and ultrasonic sensors in detecting surrounding objects and generating a composite view around a vehicle. They’re particularly useful in bad weather conditions, capable of working in fog, snow, rain and darkness that otherwise would compromise video cameras and LiDAR sensors. Processors accept the sensor inputs and then execute artificial-intelligence software to make all of the driving decisions.
But what else can you do with a mmWave sensor? One example is a liquid-level sensor in a tank. Many industrial, process-control, and public-service utilities applications require some form of level sensing.
Another interesting use is lighting control. The radar sensors detect people or motion, and then turn lights on in the presence of bodies or turn them off when no objects are nearby. HVAC usage could also benefit from this approach. No doubt, people-sensing results in considerable energy savings in buildings and parking lots, and along selected streets.
Robots and drones are good candidates for radars, too. Some manufacturing robots need to determine range, speed, and motion. The result is intelligent timing and positioning of robotic arms in industrial automated plants and other applications. Military surveillance and weapon robots come to mind as well. In addition, drones can be made safer with radars to prevent collisions and to measure range or altitude.
Security systems can also benefit, since radars detect motion at a distance and provide object detection in bad weather that may fool video cameras or IR sensors. It’s safe to say you will eventually see radars in motorcycles and bicycles. Automatic gate or garage door openers are other possibilities. It’s not hard to imagine a host of other uses once you get past the mental, technical, and price barriers of these sensors.
Safer, More Efficient Cities via Smart mmWave Sensors
One excellent application for wwWave sensors concerns traffic monitoring and control. Many, if not most, midsize to large cities suffer from traffic congestion that not only wastes time and fuel, but also adds to environmental pollution. The way to mitigate this problem is through better timing and sequence control of traffic lights at intersections and access along major highways. This requires fine-grained sensing of vehicles to determine their positions, speeds, turn intentions and directions. With this information, traffic-light timing can be adjusted to move traffic faster and more efficiently.
The inherent characteristics of mmWave sensors make them ideal detectors for traffic systems. With FMCW radar, it’s easy to determine range, speed, and angle of a vehicle. Strategically placed sensors in overhead poles, signs, or other structures can identify individual vehicles and their movements. The narrow field of view provided by phased-array antennas allows the traffic systems to monitor individual lanes (Fig. 3).
3. Radars mounted above the stop bar at intersections have a narrow range of view to count cars per lane and determine their speed.
Other benefits of a radar sensor include long-range detection up to 250 meters or short-range detection down to 5 cm or less. As mentioned earlier, these sensors also work in almost all environmental conditions, unlike video camera sensors. With appropriate computing power, they can count cars and determine their range and speed up to 300 kph (187.5 mph). And radar sensors don’t require digging up the roadway, as is the case with embedded inductive coils used at many intersections.
As cities become smarter, traffic engineers are improving the efficiency of intelligent transport systems to alleviate this growing traffic problem. With mmWave radar sensors, engineers can build next-generation intelligent traffic-monitoring systems. Radar sensors at intersections can help manage traffic lights for stops, left turns, and pedestrian crossings, speeding them up to minimize backups.
What Sensors Are on the Market?
Texas Instruments offers a complete line of mmWave radar chips that can form the basis of intelligent traffic-management systems. Available ICs include the AWR1243, AWR1443, and AWr1642. These CMOS devices, which operate in the 76- to 81-GHz range, use FMCW.
A key feature is measurement precision. For example, a highly linear, closed-loop phase-locked loop (PLL) generates the chirp ramp up in frequency to ensure greater accuracy and resolution in range measurement. Another benefit is that CMOS devices consume less power than SiGe equivalents.
These devices feature two or three transmitters and four receivers. RF bandwidth is 4 GHz, and received sampling rate is either 12.5 or 37.5 Msamples/s. The AWR1443 and AWR1642 incorporate a 200-MHz ARM Cortex-R4F processor. The AWR1443 contains a radar hardware fast Fourier transform (FFT) accelerator. And the AWR1642 has an embedded TI C674x 600-MHz DSP to handle FFTs and other advanced algorithms. Typical interfaces include SPI, CAN, CAN-FD, UART, I2C, and MIPI CS12, depending on the model.
To aid in development, TI makes available evaluation modules, reference designs, and software-development kits. Its mmWave Studio is a set of offline tools for analysis and algorithm development.