SmartFactory KL e.V Plant Scenarios for the STAR project

In the STAR project, we try to bring the safety zone detection concept to the dynamic environment. In the current scenario of the SmartFactoryKL testbed, we utilize the robot’s LiDAR sensor for its safety through movement. In the as-is scenario, there is no prediction anticipated. Hence, it does not reach the dynamic layout changes. When the obstacle is close enough to the head side of the Robot, it will stop further movement.

The STAR project brings the detection and prediction phase in the safety zone detection. To achieve this, three use cases were driven in the STAR project:

1. Human intention recognition.
2. Robot reconfiguration is based on the dynamic layout.
3. Dynamic path planning using both first and second use cases.

In the figure below you can see the diagram of these three use cases insertion into the as-is scenario.

The first use case plans to detect human activities and predict their next actions. For this matter, DFKI initiates typical workflows as the workers’ scenarios, happening during normal daily work. The behaviors of more than 10 participants were recorded, who were supposed to follow the same or similar flows in the free order. The recordings were made using wrist sensors its data are then analyzed in detail to detect the activities the humans are currently performing.

The second use case is to dynamically update the navigation route of the mobile robot, by considering human and/or other (non-)moving objects in the environment. This use case will also enable easier reconfiguration of the robot in case the layout of the environment (including the production stations) changes. The layout is actively monitored by the cameras installed stationary, and humans, as well as the objects in the layout, are detected. In case of any change, the new coordinates of the stations, where the robot should navigate, are updated.

For the third use case, these two use cases are going to be combined to have a safe environment for the workers and the hardware equipment. The newly received coordinates of the stations will be used to set the robot’s destinations. The speed of the robot and the objects in the layout will also be considered to create a collision-free navigation path for the robot. The human’s current and next activity is also one of the important aspects to take into account during the decision. In consequence, the workers’ behavior prediction from the first use case is taken into account here. Here, the STAR project will bring an indispensable functionality and aspect to the industrial environments for the safety zone detection and undemanding reconfiguration of the layout.