Autonomous Measurement and Efficient Storage of Industrial Robot Motion Data

• Applicant: Professor Dr.-Ing. Bernd Kuhlenkötter
• Subject Area: Automation, Mechatronics, Control Systems, Intelligent Technical Systems, Robotics ; Production Automation and Assembly Technology
• Term: since 2022
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 515675259

The range of applications for industrial robots is manifold: Industrial robots are used for workpiece handling, as measuring equipment or for manufacturing. Especially in manufacturing (e.g. welding, milling, painting), the robot's motion has a direct influence on the process result. Due to the design and control characteristics of the robot, a realized motion path generally deviates from the planned path. Therefore, a detailed knowledge of the motion deviations is relevant for the realization, improvement and testing of robot-based production processes.Modern measuring systems enable the detection of the motion paths of industrial robots, which opens up a wide range of analysis possibilities to investigate or improve the suitability for certain processes. In practice, only specific measurements are usually carried out in a limited number of cases, and the measurement data on the robot motion behavior is usually not available. Even existing data, e.g. from the data sheets of robot manufacturers, can only be used to a limited extent, since no detailed information about their measurement is known. Therefore, for each individual application, a large number of measurements has to be carried out, which is very time-consuming and cost-intensive. There are no methods to support such measurement processes with automated systems that go beyond the measurement of individual standardized tests (e.g. according to ISO 9283). There are also no general methods for path evaluation. For this purpose, an allocation method for motion deviations - based on the Dynamic Time Warping method and developed in preliminary work - is to be adopted and further developed.By means of available optical measuring systems, an automatic measurement of robot motions is basically possible. Due to the large number of possible settings and boundary conditions for the robot motion, however, the number of possible experiments is theoretically infinite. Therefore, a measurement method needs to be developed, in which the individual test runs are planned and carried out automatically. The system is to decide independently on the basis of previous experiments, in which parameter space further tests are appropriate to obtain necessary information. Since individual specific parameters for describing the robot motion behaviour can only cover a limited range of application scenarios, the robot measurement data are to be stored in a database system. Within the scope of the project, the basis for a measurement database is thus to be created in which the recorded robot motion data are stored. The information contained in the database provides a comprehensive description of the robot motion behaviour. With a detailed database, which can always be extended by new measurement data, a wide range of application scenarios will be supported, in which time-consuming and cost-intensive individual measurements can be avoided.

• DFG Programme: Research Grants