Towards intelligent robotic assembly: trends in industry

Citation

Chen, H., Fuhlbrigge, T.A. and Zhang, G. (2007), "Towards intelligent robotic assembly: trends in industry", Assembly Automation, Vol. 27 No. 3. https://doi.org/10.1108/aa.2007.03327caa.002

Towards intelligent robotic assembly: trends in industry

Assembly tasks using industrial robots have increased in both number and complexity over the years because of the increasing requirements of product quality and quantity. However, assembly robots are still a small portion of total robot sales each year. One of the main reasons is that it is difficult for conventional industrial robots to adjust to any sort of change. Therefore, more intelligent industrial robotic systems are rapidly expanding the realms of possibility in assembly applications. The intelligent robotic assembly system can perform assembly tasks with high autonomy and adaptability to the environments. However, the implementation of the intelligent robotic assembly system is still in its infancy. To develop an intelligent robotic assembly system, the following aspects should be considered depending on the industrial applications.

Sensor integration

The intelligent robotic assembly system should be able to deal with the semi-structured or unstructured environments. Traditional assembly systems require accurate part location devices, such as fixtures and part feeding devices, etc. to keep work pieces within tight tolerances. With the development of vision guidance systems, the intelligent robotic assembly system can take position/orientation changes to perform the assembly. For tight tolerance assembly, such as piston installation and high precision valve assembly, the vision system alone cannot guide the assembly successfully because a small position error from the vision system or part orientation change will cause failure of the assembly. In these cases, force control is necessary for assembly. Hence, intelligent robotic assembly with sensor integration will have more and more applications in industry. Recently, robot manufacture's, such as ABB, Adept and FANUC, have put more and more effort in the research and development of intelligent industrial robots. For example, FANUC has intelligent robots (www.fanucrobotics.com/) with force and vision sensors. ABB has also developed a high precision assembly system based on vision and force sensor integration (Chen et al., 2007). Other sensors, such as tactile, temperature and pressure, could be integrated into the intelligent robotic assembly system. This will enable it to carry out complicated assembly tasks that require greater dexterity:

Safety

The application of industrial robots removes humans from tedious, dangerous or harmful operations. However, robots themselves can also create dangerous conditions and threaten human safety. The intelligent robotic assembly system should be able to keep the workplace safe, especially for some complicated assemblies that need human involvement. For example, both KUKA's Safe Robot (www.kuka.com/en/pressevents/news/NN_051017_01_Teamworking.htm) and ABB's SafeMove (Kock, 2006). Both systems allow direct interaction between humans and robots. Thus, human and robot can work closely and safely.

Fixture/gripper design

Although the intelligent robotic assembly system can work in semi-structured or unstructured environments, fixtures and grippers are still needed for transportation and handling of parts. In current automotive manufacturing, the assembly line, including fixtures and grippers, has to be redesigned and installed when a new vehicle model is introduced. This requires considerable investment and lead time. Therefore, the fixtures and grippers in the intelligent robotic assembly system should be easily reconfigurable or reusable for different models or parts. One of the examples is the flexible fixture system based on modular robot cells (Negre and Legeleux, 2006).

Easy setup, installation, diagnostic and maintenance

To setup and install an assembly system currently needs a lot work and the cost is very high. Also the diagnosis and maintenance are difficult. The whole assembly system has to be shut down sometimes, which results in decreased productivity. The intelligent robotic assembly system should be easily setup, installed and maintained. Also some intelligent diagnosis should be performed.

Easy programming

To make the robot carry out an assembly task as intended even in a structured environment, usually requires an operator to “teach” or “program” the robot manually. This necessary step is mainly due to the fact that the robot lacks a human's understanding of an assembly task and the human's ease in identifying key surfaces. The intelligent assembly robotic system should be able to follow simple instructions from operators and generate the robot programs automatically. In this way, the system is able to interactively communicate with operators to perform an assembly task.

Optimal processing and minimal cycle time

Optimal processing can improve the quality of the final product. In manufacturing, minimal cycle time means high efficiency and low cost. Therefore, the intelligent robotic assembly should be able to optimize the assembly process and cycle time.

Multi robot cooperation

As the assembly tasks become more and more complicated to adapt to the rapid change in manufacturing, multiple robots cooperating with each other to finish a certain task will become more and more common. Therefore, the intelligent robotic assembly should be able to perform cooperation without too much human intervention.

Learning

Current robotic assembly systems can only repeat the programmed assembly without any improvement. Since, the systems are running continuously and a lot of data can be obtained during the assembly process, the intelligent robotic assembly system should be able to learn from the collected data and use it to improve the assembly process and optimize the assembly parameters, especially cycle time.

Other issues

There are some other issues, such as assembly for various order sizes, system cost, order execution, better inventory and order tracking, which should be considered in the intelligent robotic assembly system as well.

Because its high autonomy and adaptability to the environments, the intelligent robotic assembly system will attract more and more attention. Although some work has been done towards intelligent robotic assembly, further research and development is still needed before the intelligent robotic assembly can be used widely in industry.

Heping Chen, Thomas A. Fuhlbrigge and George Zhangare in Robotics and Automation Labs, ABB Cooperate Research Center, Windsor, Connecticut, USA