The navigation and guidance of AGV refers to the process in which AGV controls speed and steering angle based on path offset, thereby ensuring that AGV accurately travels to the position and heading of the target point. It mainly involves three major technical points:
1. Positioning
Positioning is the most basic step in AGV navigation and guidance, which determines the position and heading of AGV relative to global coordinates in the working environment.
2. Environmental perception and modeling
In order to achieve autonomous movement of AGV, it is necessary to identify various environmental information based on multiple sensors, such as road boundaries, ground conditions, obstacles, etc. AGV determines the reachable and unreachable regions in the forward direction through environmental perception, determines the relative position in the environment, and predicts the motion of dynamic obstacles, thereby providing a basis for local path planning.
3. Path planning
According to the degree to which AGVs master environmental information, they can be divided into two types: one is global path planning based on known environmental information, and the other is local path planning based on sensor information. The latter environment is unknown or partially unknown, that is, the size, shape, and position of obstacles must be obtained through sensors.
Comparison of AGV navigation methods
Early AGVs mostly used magnetic tape or electromagnetic navigation, which had simple principles, mature technology, and low costs. However, changing or expanding the path and later maintenance were more cumbersome, and AGVs could only follow fixed routes and could not achieve intelligent avoidance, or real-time task changes through control systems.
At present, the mainstream navigation method for AGV is QR code+inertial navigation, which is relatively flexible to use and easy to lay or change the path. However, the path requires regular maintenance. If the site is complex, the QR code needs to be replaced frequently. Additionally, strict requirements are placed on the accuracy and service life of the gyroscope.
With the development of SLAM algorithm, SLAM has become the preferred advanced navigation method for many AGV manufacturers. SLAM does not require other positioning facilities, and its form and path are flexible and adaptable to various on-site environments. I believe that with the maturity of algorithms and the compression of hardware costs, SLAM will undoubtedly become the mainstream navigation method for future AGVs.
SLAM can be roughly divided into two categories: laser SLAM (2D or 3D) and visual SLAM.
Visual SLAM is currently in the stage of further development and application scenario expansion. Visual SLAM has received widespread attention due to its advantages such as large amount of information and wide applicability. However, algorithms require high processor requirements, typically requiring a desktop level CPU or even a GPU. However, AGV mostly uses embedded processors, making it difficult to apply it on a large scale on small AGV devices in a short period of time.
Laser SLAM started earlier than visual SLAM, and its theory and technology are relatively mature. Its stability and reliability have been verified, and its performance requirements for processors are much lower than visual SLAM. For example, mainstream laser SLAM can run in real-time on ordinary ARM CPUs. Currently, some AGV manufacturers have launched products based on laser SLAM navigation. Undoubtedly, laser SLAM will still be the mainstream SLAM solution for a period of time.
AGV navigation and guidance technology has been developing towards higher flexibility, higher accuracy, and stronger adaptability, and its dependence on auxiliary navigation markers is becoming lower and lower. The free path navigation method of real-time positioning and map construction, such as SLAM, is undoubtedly the future development trend. I believe that in the near future, the integration of technologies such as 5G, AI, cloud computing, IoT, and intelligent robots will bring earth shaking changes to the AGV industry, and the SLAM navigation method with higher flexibility, accuracy, and adaptability will also be more adaptable to complex and ever-changing dynamic working environments. After the joint development of multiple disciplines, there will definitely be more high-end AGV navigation technology in the future.