Embedded systems, with embedded computers at their core, represent a new technological development direction following IT network technology. Their powerful and flexible applicability has gained widespread recognition in industries such as computer, communication, and information, and they have been widely applied in fields such as industrial control, traffic management, information appliances, home intelligent management systems, networks and e-commerce, environmental monitoring, and robot control [1]. The emergence and development of embedded systems will truly realize the "ubiquitousness" of computers.
The development of robotics technology has always been closely linked to the development of embedded systems. Research in robotics is essentially the application of embedded technology, and the development of embedded technology inevitably promotes the level of robot intelligence. Since the mid-1970s, due to the development of intelligent control theory and the emergence of microprocessors, robotics has gradually become a research hotspot and has achieved significant progress. Currently, embedded systems are widely used in robot control systems.
1. Embedded systems
1.1 Definition and characteristics of embedded systems
From an application perspective, embedded systems are a branch of computer development that emerged after the 1970s. They are application-centric, based on computer technology, and feature customizable hardware and software. They are suitable for application systems with stringent requirements regarding functionality, reliability, cost, size, and power consumption. Simply put, an embedded system is a dedicated computer system embedded into a target system architecture.
An embedded system typically consists of an embedded microprocessor, peripheral hardware devices, an embedded software operating system, and user applications.
The program consists of four main parts, used to implement functions such as controlling other external devices and exchanging data over a network. For specific application requirements, the embedded system is embedded into the target hardware and software architecture.
Embedded systems have the following main characteristics compared to ordinary PC systems:
(1) Embedded systems have low power consumption, small size, and strong specialization. Embedded CPUs work in systems designed for specific user groups and can integrate the tasks performed by many boards in a PC into the chip, which is conducive to the miniaturization of embedded system design.
(2) In embedded systems, the software is generally embedded in the memory chip or the microcontroller itself to improve execution speed and system reliability. Both hardware and software must be designed efficiently, the system must be streamlined, and the quality of the software code is very important. The operating system is generally integrated with the software.
(3) Embedded system development requires specialized development tools and development environment.
2. Application of Embedded Systems in Robotics
Embedded controllers are becoming increasingly miniaturized and functional. Micro-robots and special-purpose robots are also gaining greater development opportunities, with significant improvements in both control system structure and robot intelligence. Intelligent robotic pets, exemplified by Sony's robotic dog, are a typical example of embedded robot control systems. Besides performing complex motion functions, they also possess advanced human-computer interaction capabilities such as image recognition and voice processing, and can mimic animal expressions and movements. The Mars rover is another prime example; this highly technologically advanced mobile robot, valued at $1 billion, uses the VxWorks operating system and can operate autonomously without contact with Earth. The following analysis examines the application of embedded systems in robotics from three aspects: motion control systems, remote control, and video monitoring systems.
2.1 Motion Control System
The motion control part of the robot is generally completed using ARM7, mainly because the whole system has high real-time requirements, and using ARM7 to specifically control the servo can better meet the requirements.
Figure 2.1 Block diagram of motor control implemented with ARM7 Figure 2.2 Framework diagram of remote control system
Figure 2.1 is a block diagram of motor control implemented by ARM7. The ARM7 receives data via serial port and parses the received data according to the defined serial communication protocol to obtain the direction of rotation and number of revolutions for each motor, thereby controlling the motor's rotation. Serial data reception is achieved through interrupts. Once data arrives, an interrupt is generated. In the interrupt service routine, the newly sent data is saved, and a flag is set to true to notify the master.
When new data arrives, the motor driver program can be called to control the motor's movement.
2.2 Remote Control
Figure 2.2 is a framework diagram of the remote control system. Any user with internet access can remotely simulate the robot's motion trajectory after receiving the robot control commands input by the user. The simulation will then perform inverse kinematics on the robot's motion trajectory to obtain the motion data of each control joint, i.e., the corresponding motor. This data is transmitted to the ARM9 control board at the near-end control center via the internet. It is then forwarded to the real-time control board ARM7. ARM7 then controls the servo driver to make the motor move along the predetermined trajectory, thereby realizing the remote control of the robot.
A crucial function of the remote user terminal is the ability to view the robot's motion posture in real time, making a video client essential. Based on a server/client model, a video server runs on the ARM9 control board. This video server connects to a camera with a USB interface. The camera captures the parallel robot's motion status in real time and encodes the captured images. The ARM9 control board then transmits the encoded and compressed image data to the remote client via Ethernet. Upon receiving the image data, the remote client decodes and displays it to form a video image, allowing the user to observe the robot's motion.
The ARM9 control board is the core of the entire system, acting as its data center and control hub. On one hand, it encodes the video data captured by the camera and sends it to the remote user terminal via Ethernet; on the other hand, it is also responsible for parsing and forwarding control commands from the remote user terminal to the ARM7 control board and receiving feedback information.
Figure 2.3 Remote Machine Command Flow
Figure 2.3 illustrates the data flow of remote commands and the data flow of robot motion status feedback to the remote user. The remote user inputs the required robot motion posture and position commands into the remote PC via the human-machine interface. These commands are transmitted over the network to the ARM9 control center, which acts as the local PC in the local debugging mode. After receiving the remote data, the ARM9 control center processes the data and then sends the serial data packets to the ARM7 control board. The ARM7 control board receives the data, parses it, and drives the servo drivers of the motors, thereby achieving robot control.
Due to the accuracy requirements of control commands and the actual network conditions, the TCP protocol is currently the most commonly used protocol for transmitting control commands.
2.3 Video Surveillance System
Figure 2.4 shows the overall framework of the video surveillance system. The camera with the USB interface is connected to the embedded development board with ARM9 (uclinux operating system). The Ethernet controller of the development board is connected to the router. The router is connected to Interact. The monitoring platform is a PC connected to the Internet at a remote location.
Figure 2.4 System Structure Diagram
Image data is captured by a camera. The ARM9 reads the data from its buffer via a USB interface and compresses and encodes the image data. Then, it transmits this data using socket programming, acting as a server. Once a client (monitoring platform) connects, it sends the compressed data. The monitoring platform receives the data, decodes it, and displays it on its terminal. Controlling the number of image frames displayed per unit time creates a smooth animation effect.
3. Development Trends of Embedded Systems in Robot Applications
With the increasing maturity of network technology, embedded products that support network functions will be used more and more widely. This will not only become a hot topic in the development of embedded systems, but also a research hotspot in robotics. The future research on robotics based on embedded systems has the following trends:
(1) High integration, low power consumption, and miniaturization. With the miniaturization and intelligence of embedded systems, robots will develop towards high intelligence, high integration, and miniaturization;
(2) Provide a convenient and fast human-machine interface. Embedded systems are widely used mainly because of the convenience and speed of their multimedia human-machine interface, which makes the operation of robots more convenient, faster, and more user-friendly;
(3) Remote control will inevitably become a trend.
4. Summary
Embedded systems play a crucial role in robot control systems, particularly in motion control video image acquisition, transmission, display, and monitoring. They are key to ensuring the real-time performance and accuracy of the entire control system. With the support of embedded and multimedia network technologies, remote control and monitoring of robots will become a reality. Furthermore, with the further development of embedded systems and multimedia network technologies, robot technology will have even broader prospects for development.
