Offline robot programming assisted by task demonstration: an AutomationML interoperable solution for glass adhesive application and welding

Overview

• This paper introduces a new robot programming system that aims to make it easier for non-experts to program robots for manufacturing tasks.
• The system captures manufacturing skills from task demonstrations performed by skilled workers and uses this information to generate robot programs.
• The researchers demonstrate the system's effectiveness in two manufacturing processes: glass adhesive application and welding.

Plain English Explanation

The researchers have developed a new way to program robots for manufacturing tasks that is more intuitive and accessible to non-experts. Typically, programming robots requires expertise in both robotics and the specific manufacturing process, which limits their use to a wider range of potential users.

This new system works by capturing the manufacturing skills of skilled workers as they demonstrate a task. It uses a magnetic tracking system like the one described in this paper to record the orientation and velocity of the worker's movements, as well as positional data from CAD/CAM software.

The system then translates this data into robot-friendly instructions, including the correct robot poses, orientations, and velocities needed to perform the task. This integrated approach to capturing both robot and manufacturing data is similar to the SmartFlow system.

The researchers tested the system on two common manufacturing processes - glass adhesive application and welding - and found that the robot paths generated were within the functional tolerance range, meaning the robot could perform the tasks accurately.

Technical Explanation

The key innovation of this work is the development of an intuitive robot programming system that can capture relevant manufacturing skills from task demonstrations by skilled workers. This is in contrast to traditional offline robot programming, where the programs often lack critical parameters like robot orientation and velocity that are necessary for executing specific manufacturing tasks.

The system uses a magnetic tracking system to record the orientation and velocity of the worker's movements during a demonstration, while positional data is extracted from CAD/CAM software. This data is then transformed into robot-friendly Cartesian space and validated in simulation before being used to generate the final robot program.

The researchers developed a new syntax called PathML, which is based on the AutomationML standard, to integrate the robot and manufacturing data across the different elements and stages of the manufacturing system. This allows for a more cohesive and transferable representation of the required skills.

The effectiveness of the system was demonstrated through experiments on two common manufacturing processes: glass adhesive application and welding. The results showed that the robot paths generated by the system were within the functional tolerance range, indicating that the robots could accurately perform the tasks.

Critical Analysis

The paper provides a promising approach to making robot programming more accessible to non-experts, but there are a few potential limitations and areas for further research:

1. The system relies on skilled workers to provide the initial task demonstrations, which may not always be available or practical. Exploring ways to generate these demonstrations automatically or semi-automatically could further improve the system's accessibility.

2. The experiments were conducted on relatively simple manufacturing tasks, and it's unclear how well the system would scale to more complex processes. Further testing and evaluation on a wider range of manufacturing applications would be valuable.

3. The paper does not address the potential for errors or inconsistencies in the worker's demonstrations, and how the system might handle or mitigate these issues. Exploring robust techniques for handling variability in the demonstration data could enhance the system's reliability.

Overall, the research demonstrates a promising approach to making robot programming more accessible, but additional work may be needed to fully realize its potential and address some of the identified limitations.

Conclusion

This paper presents an innovative robot programming system that aims to make it easier for non-experts to program robots for manufacturing tasks. By capturing the manufacturing skills of skilled workers through task demonstrations, the system can generate robot programs that accurately execute the required motions and processes.

The experiments conducted on glass adhesive application and welding processes showed promising results, with the robot paths generated by the system falling within the functional tolerance range. This suggests that the system could be a valuable tool for expanding the accessibility of robotics in manufacturing, potentially opening up new opportunities for automation and improving productivity.

Further research is needed to address some of the identified limitations, such as the reliance on skilled worker demonstrations and the need for more extensive testing on complex manufacturing tasks. However, the core ideas presented in this work represent an important step forward in making robot programming more intuitive and accessible to a wider range of users.