Millimeter-wave (mmWave) radar has found a home in many areas from self-driving cars and aviation applications, but low-cost radar systems are equally useful in a wide range of embedded applications. They can be used to detect people within a car to controlling entry doors more accurately than other methods such as ultrasonics.
I had a chance to check out Texas Instruments (TI) IWR6843 60-GHz mmWave development kit (Fig. 1) that targets industrial applications. TI has other kits that deal with different frequencies and applications, such as automotive. The chip, with its three transmitters and four receivers, uses a C674x DSP to handle signal processing. An ARM Cortex-R4F with ROM firmware handles object detection and interface control. Communication is possible with CAN-FD, UARTs, SPI, and I2C. It also maintains six ADC channels for other chores. Internal memory features ECC protection.
1. The IWR6843ISK is an antenna plug-in board for Texas Instruments’ mmWave development kit.
What was equally impressive were the applications and application notes for the kit that are part of the mmWave Studio. The Studio is designed to present the mmWave sensor functionality through demos and system tools like the web-based mmWave Demo Visualizer and Sensing Estimators. The web-based tools connect to the hardware via a serial port, making it possible to be up and running in a very short period of time. The visualizer allows the configuration to be saved so that it can be used in an application.
Application development is done using TI’s Code Composer Studio and the mmWave SDK. I didn’t get too far with that yet due to time, but there are a number of projects to get started with from using the hardware with ROS (robot operating system)-based robots to gesture control.
The platform has many advantages over alternative range information systems, including the ability to detect and see through materials like glass and drywall. It can essentially ignore environmental factors such as smoke, rain, and lack of light. The system can generate a point cloud that’s sufficient to distinguish people and other objects accurately.
Though the kit I tested uses a PCB-based antenna, developers can also take advantage of the antenna-on-package (AOP) solutions (Fig. 2). It uses the same system-on-chip (SoC) I tested with antennas mounted on top of the package. This presents a number of advantages in addition to greatly simplifying implementation. In particular, all of the RF design and certification is handled by TI; therefore, a developer can use the AOP solution without having to address these issues.
2. The chip on the left is the IWR6843 antenna-on-package (AOP) that includes electronics and antennas for a complete millimeter-wave solution.
There are benefits to using larger custom antenna arrays and often systems will need to employ multiple chips to handle such arrays. Nonetheless, a simpler, more compact solution is suitable for many applications.
TI has taken a lot of the work out of developing an embedded radar solution—to the point that prototypes could be up and running in a very short period of time. They also provide direction and movement information that isn’t possible with simpler proximity sensors, as well as operate in environments that would stifle or completely curb technologies like visual imaging systems. Checking out mmWave solutions is inexpensive and quick, so it’s definitely worth looking into.