Industrial sites still use video cameras to monitor production, the same cameras that have been around industries for a long time. Almost all control rooms are covered by screens that show different parts of the production and activities combined with other measurement data. It is an overload of information and types of data provided to the operator in charge. Smart video and AI are not adopted as much as in other industries - the need for robustness sets high requirements on the technical solutions, especially when introducing something new.
The lidar sensor is constantly scanning its environment with thousands of measurements that create a 3D model. This allows analysis in terms of volumes, shapes, and distances, compared to video analysis for example. A volume change can be an increase or loss of volume, and it is also possible to filter objects based on the volume. A volume increase in a static scene means that new objects enter the scene and can be tracked. This feature supports tracking of the most common objects such as people and vehicles. Smaller disturbances - such as birds, smaller animals, and vegetation can be filtered and removed. Spillage and other production deviations can also cause a volume increase if the area should be empty normally. A volume loss is also serious, it can be caused by a breakdown or collapse in a mountain or the infrastructure. Volume change analysis can be used in many ways to separate changes to the desired situation. It is also possible to have different rule for different areas, such as vehicle roads and safe walking paths for people.
An increase in the volume is detected as an object that can be sorted out as a person, vehicle, etc. based on the size and shape. When tracking an object, for example, a person, we can follow the person’s walking path and velocity. The velocity parameter can also be used to identify dangerous situations of moving objects in the same environment. Since the lidar sensor sees in 3D, the distances between objects become natural and accurate. The safety of a site can be improved when every moving object can be visible and followed without breaching integrity. Instead of equipping all vehicle with sensors and people with wearables, fewer sensors can monitor the entire vehicles fleet and all activities in the environment and alarm if potential danger is detected, e.g. collisions.