Research and development of AGVs in China started relatively late, but it has a history of over 40 years since the first AGV came out. During this period, navigation technology has always been the most critical point for AGVs, as the navigation method largely determines its efficiency and flexibility. With increasing demands for complex scenarios, single navigation method can no longer satisfy the needs, and integrated navigation has become a breakthrough to solve the issues.
Currently, navigation methods in the market maily include:
Electromagnetic guidance is a relatively traditional guidance method. The implementation form is to embed metal wire on the driving path of AGV, and load low-frequency and low-voltage current on the metal wire to generate magnetic field. The navigation is realized by identifying and tracking the strength of the guiding magnetic field through the on-board electromagnetic sensor, and the task is completely specified by reading the embedded RFID card.
The main advantages of electromagnetic guidance are that the metal wire is buried in the ground with strong concealment and is not easy to damage. The guiding principle is simple and reliable, has no interference with sound and light, and its constructing cost is low. However, the laying of metal wire is inconvenient, it is difficult to change or expand the path, and it is easily affected by ferromagnetic materials such as metals.
Electromagnetic guidance is widely used in manufacturing industries (such as automobile manufacturing) where running circuit is relatively simple and requires 24 hours of continuous operation.
2. Magnetic Tape Guidance
The principle of magnetic tape guidance is similar to that of electromagnetic guidance, which is laid on the formal path of AGV, and the guidance mode is realized through the recognition of magnetic field signal by on-board electromagnetic sensor.
The main advantages of magnetic tape guidance are that the technology is mature and reliable, the cost is low, the laying of tape is easier, the expansion or change of path is easier than electromagnetic guidance, the running circuit is obvious, and there is no interference to sound and light. But with an exposed path, it’s vulnerable to mechanical damage and pollution, requiring regular maintenance, and vulnerable to the influence of ferromagnetic materials such as metals. Once AGV performs a task, it can only move along the fixed tape and cannot change the task.
Magnetic tape guidance is suitable for ground-embedded, light-load traction, and can be used in non-metallic ground and non-demagnetized indoor environments. The operation can be stable and lasting.
3. QR Code Guidance
The marking of the coordinates is realized by a QR code on the ground. QR code guidance is similar to magnetic nail guidance, but the coordinate markers are different. The automatic guided vehicle scans the QR code on the ground through a camera, and obtain the current position information by analyzing the code information. QR code navigation is usually combined with inertial navigation to achieve precise positioning.
QR code navigation is currently the main stream in the market. The main reason is that Amazon acquired the KIVA QR code navigation robot at a high price, whose chessboard-like working mode is impressive. Since then domestic e-commerce and smart warehouses have adopted QR code navigation robots. The stand-alone cost of a mobile robot guided by QR code is low, but numerous QR codes need to be laid on the project site, which is easy to wear, therefore brings a high maintenance cost.
4. Laser Navigation
Laser navigation normally refers to laser guidance based on reflector positioning, of which principle is to install reflector with accurate position around AGV driving path with laser scanner installed onto the AGV. The laser scanner sends out a laser beam when AGV is walking, which will be directly reflected back by multiple sets of reflection plates laid along the AGV travel path. The trigger controller records the angle when the rotating laser head meets the reflector. According to these angle values and the actual position of the reflector, the controller calculates the absolute coordinates of AGV. Based on this principle, very accurate laser guidance can be realized.
The laser navigation method enables AGVs to plan the route flexibly, locate accurately, and drive flexibly and changeably. The implementation is quite convenient and it can adapt to various practical environments. As reflectors are installed in a high place, it is not easy to damage them. However, reflectors cannot be shielded during operation, otherwise the positioning result will be affected. With high cost, laser navigation is not very popular in AGV market, but it is a more advanced technology which is perceived as a navigation type that will gradually replace some of the traditional navigation guidance methods.
5. Vision-based Navigation
Vision navigation is to obtain the image information around the running area by the vehicle-mounted visual sensors to realize AGV navigation. For the hardware part, we will need a down-looking camera, supplemental lighting and lens hoods to support the realization of this navigation mode. We can use rich ground texture information, and calculate the displacement and rotation between the two images based on phase correlation method, and then obtain the current position through integral calculus.
In this method, the ground texture is captured by the camera during the moving process of the mobile robots to build map automatically, and then the ground texture information obtained in the running process is registered and compared with the texture image in the self-built map, so as to estimate the current posture of the mobile robot and realize the positioning.
Vision navigation AGV (VGV) is rarely used in the market. Its advantages are low hardware cost and accurate positioning. However, its disadvantage is that there are needs of texture information from the ground, and when the operation site is large, the time of drawing navigation map is longer.
Single method navigation mode can adapt to a limited number of scenes, and the multi navigation integration mode can make AGV have more than two navigation characteristics and meet the needs of more complex scenes. With the development of intelligent warehouse, we believe that integrated navigation will become a trend.