“In my last blog post, I discussed how Texas Instruments (TI) millimeter wave (mmWave) sensors can provide edge intelligence for robotic arms in factories. Now, I want to discuss how millimeter wave technology provides edge intelligence for autonomous robots, enabling sensors to make real-time decisions to slow down or stop the robot and ensure its continued performance in industrial robot applications.
In my last blog post, I discussed how Texas Instruments (TI) millimeter wave (mmWave) sensors can provide edge intelligence for robotic arms in factories. Now, I want to discuss how millimeter wave technology provides edge intelligence for autonomous robots, enabling sensors to make real-time decisions to slow down or stop the robot and ensure its continued performance in industrial robot applications.
TI millimeter wave sensors can be used in systems designed to help industrial robots avoid collisions and solve the key problem of collaborative interaction with robots that interact with humans and other objects. If the edge requires additional machine learning processing, millimeter-wave sensors can work seamlessly with industrial-grade processors (such as Sitara™ processors) to provide additional intelligence.
Just as advanced driver assistance systems in automobiles can use millimeter waves for surround panoramic monitoring and obstacle detection, millimeter wave sensors can help solve similar challenges for automated guided vehicles (such as logistics robots, delivery vehicles, forklifts, and hydraulic vehicles) ,As shown in Figure 1.
Picture 1: Various automatic guided vehicles
3D point cloud detection
The millimeter wave sensor with three-transmitter and four-receiver antenna configuration can use angle information on the azimuth and elevation planes to achieve 3D object detection up to 30 m in order to detect objects from a height. This is extremely useful for vehicles such as forklifts, because their sensors are located high above the ground. A single sensor can detect objects within a 120-degree field of view, thereby minimizing the number of sensors required to surround the monitoring system.
High resolution and accurate detection
Since the millimeter wave sensor works at a bandwidth of 4 GHz, it can recognize individual objects separated by 4 cm and operate in areas that may be blocked by darkness, dust, or physical barriers. This high-resolution capability is necessary so that the sensor can accurately count and identify objects or people, and trigger appropriate actions in real time, such as stopping the robot before a collision.
In addition to real-time object sensing and collision avoidance, millimeter wave sensors also provide additional functions to make smart industrial robots possible.
Ground speed and edge detection
TI’s millimeter wave technology provides sub-millimeter accuracy and high resolution through Doppler frequency shift, and realizes accurate ground speed sensing, enabling the sensor system to calculate the ground speed of the vehicle and detect the edge of the ground, such as the loading and unloading platform where the wheels may slip, To avoid unrecoverable situations.
Transparent object detection
TI millimeter wave sensors can detect transparent objects, such as glass and plastic, and dark objects, which may challenge some light-based technologies. Improved detection accuracy helps avoid potential accidents such as collisions with glass barriers or objects.
Comply with SIL-2 standard
TI’s 60 GHz millimeter wave sensor can help the system meet the IEC Safety Integrity Level (SIL)-2 standard for event management when human-computer interaction is close. When designing a SIL-2 certification system, engineers can directly use TI millimeter wave sensors to provide functions that may require the system to implement additional safety processor systems or redundant sensor systems to obtain certification.
TI millimeter wave sensors in autonomous robots
As part of the digital signal processor, TI’s highly integrated millimeter-wave radar sensor has advanced clustering and tracking algorithms to provide intelligent autonomy at the edge. Figure 2 shows the TI millimeter wave chip with integrated processing.
Figure 2: A typical front-end sensor solution and an integrated TI millimeter wave sensor that can realize edge intelligence.
Because TI millimeter wave sensors are not affected by the environment (rain, dust, smoke) and lighting conditions, and can be sensed by materials such as plastic, they can effectively detect objects in the robot’s path without external lenses, apertures or sensor surfaces.
More edge intelligence
Deep learning is a subset of machine learning and is becoming more and more popular in the industrial field. TI provides hardware and software support to help designers bring deep learning inference to various applications including robotics.
For applications where the smart sensor’s own data processing capacity is insufficient, it is equipped with a high-performance Arm up to 1.5 GHz®Cortex®-The Sitara Am57x processor series with A15 core and dual-core C66x processor can run deep learning inference and traditional machine vision algorithms, which can solve the machine learning needs of applications such as predictive maintenance and remaining service life, or based on sensors that exceed existing functions Enter to make a decision. Sitara AM57x processors provide dedicated hardware for industrial communications (EtherCAT, Profinet, time-sensitive networks, Profibus, Ethernet/Internet Protocol), and can be used as the central processing unit of the robot controller.
In short, the integrated millimeter-wave sensing solution for robust object detection and the Sitara processor for enhanced machine learning provide intelligent solutions for area scanners of robotic arms or collision avoidance for autonomous robots. Please refer to our resources to start development.