Electronic Enthusiasts Network reported (text / Li Ningyuan) Robots, a hot topic that can always trigger people's discussions. With the process of industrial innovation, industrial robots have always been one of the devices that have attracted much attention in the wave of industrial automation. From the current situation, the current industrial robots can be called semi-intelligent devices at most, and it will take a long time to evolve into intelligent devices. . With the popularity of robot mobile technology in recent years, service robots have also come into everyone's sight.

Whether it is an industrial robot or a service robot, the tasks they are given are nothing more than from perception to decision-making to execution. For the robot itself, the motion control of the executive layer is its most basic and core capability. Perception can enrich the information acquisition of machines by integrating sensors such as lasers and vision. How to make decisions depends on the capabilities of the upper-layer software and intelligent algorithms of each manufacturer. The final implementation of the operation control at the execution layer still depends on the "motility" given to it by the internal control elements of the robot itself.

At present, MCUs that are longer than control are still the first choice for robot manufacturers. DSP and FPGA also have many applications, especially in the field of machine vision. DSP is very popular, but MCU is still the most widely used choice in the robot execution layer. After all, about 80% of the controllers built in a single robot arm are MCUs. , the rest are high-efficiency processing components such as MPU and DSP; the important mobile functions of service robots also depend on precise MCU motor control at the execution level.

For MCUs for robot applications, in addition to providing easy-to-develop embedded platforms and design tools, it is also very important to establish a complete communication environment around the MCU to handle various industrial communication protocols such as USB and SPI. The improvement of MCU computing performance is also very important to meet the needs of industrial robot precision improvement. Here is a brief list of some MCUs used in robots to see how they develop and play in the robot market.

MCU integrated ROS framework

MCU integrated ROS can see Renesas. RA MCUs are widely used in many fields. Recently, they have integrated their MCUs with ROS (Robot Operating System) – EK-RA6M5 for the development of Industry 4.0 and Industrial Internet of Things to deal with robot applications. RA's RA6M5 is used to integrate ROS, which is a 32-bit MCU with Arm Cortex-M33 core.

RA6M5 itself runs the CoreMark algorithm from flash memory with low power consumption as low as 107uA/MHz and large-capacity embedded RAM, which is extremely suitable for robot applications. It can be used for Pub/Sub communication interfaces (such as Ethernet and USB) between ROS nodes. ). The RA6M5 also integrates an Ethernet MAC with dedicated DMA to ensure extremely high data throughput. The RA6M5 ecosystem is supported by FSP built on FreeRTOS, which ensures that the MCU has high scalability and can use other real-time operating systems (RTOS) and middleware.

The integrated ROS framework of the EK-RA6M5 enables standardized integration of MCUs in the ROS 2 data space, and the Renesas e2 studio integrated development environment running under Windows and Linux systems also supports ROS. The EK-RA6M5 can control multiple servos via single-wire UART and RS-485, and supply the proper voltage to the entire system.

Integrating ROS when the MCU performance is good enough, the purpose of this set of EK-RA6M5 is very clear, which is to simplify the development of robots, reduce the cost of robot development, and form an ecosystem based on the merger of RA MCU and ROS to attract robot manufacturers. This grabs the robot MCU market. This is also the approach chosen by many manufacturers. MCU is playing an ecosystem, and this time only ROS is added. Of course, the premise of playing the ecology is that the MCU performance must be good enough.

Optimizing MCUs for Robotic Applications

Of course, there are also ways to play without ROS. KE series is NXP's MCU based on ARM core in operation control. The KE series does not integrate the ROS framework for robot applications alone, it addresses the high reliability and robustness of machines in harsh electrical noise environments.

KE's EMC electromagnetic compatibility and ESD electrostatic discharge protection features are important enough for industrial grade applications. The 32-bit MCU has always been running at low voltage, and is often limited by the harsh electromagnetic environment in robot control. The most important thing of the KE series is to solve this problem, and solve the problem of MCU in robot applications.

In terms of performance, the KE1xF MCU, the high-end series of MCUs in the KE series, is actually excellent enough. It provides a 5V solution through the high-performance Arm Cortex-M4 core running at a frequency of up to 168 MHz. This family of 8-channel PWM and support with dead-time insertion allows precise machine motion control, and FlexIO also provides flexibility for serial communication interface implementation.

In addition to special optimization for application difficulties, KE has digital signal processing capabilities with floating-point units, which can provide computing power for machine control algorithms and sensor data processing. This approach does not start from a macro perspective to seize the opportunity like integrating ROS, but instead focuses on solving the outstanding problems of MCU applications in robots. Solve the problem from the problem, and ensure the competitiveness of its own MCU in the application.

write at the end

In fact, only in terms of performance, all MCU manufacturers have products that can be used for robot applications. The above two examples confirm the trend of MCU applications in the robot industry. First, when the MCU performance is good enough, in order to seize the robot market, the big manufacturers will cooperate with ecological partners to carry out ROS integration (of course, they can also do it themselves). In short, the purpose is to create a platform with a complete arsenal to attract robot designers. Enter the circle; secondly, aiming at the pain points in robot applications, supplement the MCU with special capabilities to enhance competitiveness.

Either way, MCU manufacturers hope to ride the wave of industrial automation and make a fortune in robots.

Disclaimer: This article was originally created by electronic enthusiasts, please indicate the above source for reprinting. If you want to join the group for communication, please add WeChat elecfans999, and if you need to submit an interview, please send an email to [email protected]

Edit: jq

Leave a Reply

Your email address will not be published. Required fields are marked *