General This presentation will give brief introduction of the sensors used for logistic autonomous guided vehicles. Logistic automation becomes one of the most important AI technology applications in industries. Autonomous forklifts and AGVs are mostly used moving vehicles in logistic automation application, typically in the warehouse for onsite short distance transportation. The merits of using such automatic devices include higher accuracy, higher productivity, and lower operation costs. To achieve the best perfection of the performance, multiple sensors are used on each autonomous forklifts and AGV for vehicle localization, object detection, collision avoidance, etc. However, there are many types of sensors, some of them are expensive compared with the total cost of a autonomous forklift or AGV. Additionally, the sensor selection also affects the complexity of the sensor signal processing unit’s hardware and software. Therefore the sensor selection should be considered at the system design phase. Typically an autonomous forklift or AVG sensing system is equipped with: 1) Active sensors: such as distance sensors which include 2D or 3D LiDAR, millimeter wave radar, sonar, etc., which detect the distance and reflectivity information of the surrounding area, and the processor process the information for configuring digital map and detect the geometric features nature of surrounding objects. Distance sensors are the key sensor for the autonomous vehicle, however, quite often they also take a big portion of the costs of all sensors. 2D and 3D LiDARs are the widely used distance sensors. 2) Passive sensors: such as imaging sensors which include spectral cameras, IR cameras, stereo cameras, etc. which provide detail information of the objects and allow the processor to use computer vision method to detect the surrounding area. They are cost effective but the distance information are not as precise as LiDAR’s. Comparison of the sensors can be seen in the following table: (Table) For most industrial application, 3D LiDAR is too expensive and mostly 2D LiDARs are used,which provide sufficient information for generate digital maps, navigation and collision avoidance. The requirements for the processing unit is also lower since much less data are treated. Examples of the applications: (Examples)