Composition of Robot System

Abstract

Robots are widely used in industry, medicine, agriculture, construction, and the military. A robot system is typically composed of machine, drive, control, sensing, intelligence, and multi-machine collaboration, which is collectively composed of robots, operating objects, and environment. As a teaching and research platform, a robot is ideally suited as an engineering object to learn and gain expertise in software programming, embedded technology, control technology, sensor technology, mechatronics, image processing, pattern recognition, data communication, and multi-agent system. This chapter describes the components of a robot system, including the computing platform, controller, actuator, sensors, and software composition.

Appendices

Further Reading

Analyze how to perform sensor selection and quantity configuration according to the application objectives of an autonomous vehicle system.

• Reference: Du M (2022) Autonomous vehicle technology: global exploration and Chinese practice. Springer

The environment perception task of autonomous vehicles is very complex, which can be summarized into the following four main aspects:

• 1. Road shape estimation and cognition.

• 2. Detection and identification of static obstacles, traffic participants, and other dynamic obstacles.

• 3. Vehicle state estimation and motion compensation.

• 4. Map-aided or mapless positioning.

The perception module is responsible for providing the real-world model for the path planning module, which is composed of three submodules: road environment perception, vehicle motion sensing, and information fusion.

The environment perception submodule uses optical binocular stereo camera, optical monocular camera, infrared camera, LiDAR, microwave radar, ultrasonic radar, sonar, and other sensors to obtain the three-dimensional environment information around the vehicle. On the premise of meeting the system requirements, the key task of the submodule is to reduce the input information redundancy as much as possible and to design and build the architecture optimal minimum vision and radar combined sensing system.

The vehicle motion sensing submodule mainly uses inertial navigation sensor, odometer, and GPS receiver to provide the vehicle’s azimuth, attitude angle, motion speed, and other information for global navigation. Its objective is to measure the real-time acceleration and viewpoint of the vehicle. The measured data are the acceleration and angular velocity along three axes. The main sensing devices are the accelerometer (when there is motion acceleration, an inaccurate inclination angle will be measured. That is to say, when the vehicle inclines, the accelerometer alone will not be able to get the accurate inclination angle) and the gyroscope, which measures the rotation speed, vehicle pitch angle, and roll angle (obtaining the viewpoint through continuous integration instead of the rotation viewpoint).

Since the middle of October 2017, Google’s Waymo autonomous vehicle has started open road test in Arizona. The sensor configuration of the Waymo autonomous vehicle is as follows.

LiDAR System: It includes three types of LiDAR: short-range LiDAR, which can provide 360° coherent field of vision, high-definition mid-range LiDAR, and a new generation of long-range LiDAR, which can detect objects at a distance equivalent to almost three football fields. This system functions effectively both during the day and at night.

Millimeter Wave Radar System: It uses multiple wavelengths to detect objects and motion. These radar waves can bypass objects such as raindrops, enabling the system to operate successfully in various weather conditions such as rain, snow, and fog. Moreover, the millimeter wave radar system is unaffected by nighttime conditions.

Vision (Camera) System: The vision system includes cameras designed to observe the surrounding environment like human beings, but it has a 360° synchronous field of vision, while human drivers can only see 120° of vision. Because the HD vision system can detect colors, it can help the system identify the strobe lights on traffic lights, construction areas, school buses, and ambulances. Waymo’s vision system consists of a number of high-definition cameras designed to see distant objects in the daytime and in low-light conditions.

Additional Sensors: Waymo includes many additional sensors, including an audio detection system that can hear the sirens of police cars and ambulances. Relying on the cooperation of sensors and software, Waymo can sense the surrounding world. It can recognize vehicles, pedestrians, bicycles, and obstacles and also distinguish the color of traffic lights, temporary parking signs, etc.

Exercises

• 1. Which mobile chassis drive is often used for transportation in urban scenarios, and why is the omnidirectional drive not used?

• 2. For the two-wheeled differential chassis model, if the speed of the car body is V, the speeds of the left and right wheels are V_L and V_R, respectively, the angular velocity of the car body is ω, the turning radius is R, the distance between the left and right wheels is D, and the distance of two wheels to reach the center of the car body is d, please solve its kinematic equations of vehicle body speed and vehicle body acceleration.

• 3. Briefly describe the components of a robot system from the perspective of control and the relationships among the different parts.

• 4. A motor rotates a circle to generate 12 pulses, the diameter of the wheel is 68 mm, and the reduction ratio of the motor reducer is 61.6 : 1. If the number of pulses per unit time is X, please calculate its speed according to the principle of the Hall effect.

• 5. Describe the advantages and disadvantages of LiDAR and briefly summarize the differences between single-line and multi-line LiDAR.

• 6. What are the measurement methods, advantages, and disadvantages of RGB-D cameras when measuring object depth? How does its measurement principle differ from that of stereo cameras?

• 7. Cameras are classified according to the exposure method as rolling shutter and global shutter. What are their differences and usage scenarios?

• 8. Briefly describe the difference between L2, L3, and L4 of autonomous driving.