Flexible Automation with ‘‘Seeing’‘ Robots
ISRA VISION Posted 08/13/2003
06/02/2003 -- One of the tasks that robots are required to do is to grip large body panels securely, even if their position is not precisely defined. One facet of this is 3D position recognition for robots unloading automotive side panels. Whether removing parts from the conveyor at the press plant or from the shipping crate during assembly, vision systems enable this process with great accuracy. The results are less downtime of costly machinery and a significant reduction in manual intervention.
Internal tension and robust transport of body components routinely lead to problems in the robotic handling of these parts. Two interfaces in the logistic system of automotive manufacturers are particularly trouble-prone: removal of newly pressed body parts from the conveyor and insertion of the same parts into the shipping container. The second task is removal of body parts from the shipping container and their placement on the welding fixture.
For both of these handling tasks, ISRA VISION SYSTEMS has developed, tested, and installed special object recognition systems, which provide excellent process security.
Picking up large and complex components—e.g. automobile side panels from the press conveyor belt—requires determination of the parts’ position and orientation while the belt is moving. The robot can use this data to pick up the part securely and independently of deviations from the standard position. This also when works when multiple presses are serviced. Using a simple ‘‘teach’‘ process, the geometrical properties and relevant coordinates of different parts can be stored.
Determination of Reference Points in Three Dimensions
The removal of body components from the transport container often fails in conventional plants because parts are not precisely positioned in their fixture.
Through transportation or other factors, individual parts may jump out of their holders. Their position then differs in three dimensions from the standard position. In cases such as this, ‘‘blind’‘ robots are unable to pick them up, even with sophisticated pickup techniques. The only remedy up until now has been time-intensive manual intervention.
When the ISRA vision system is used, the failure never occurs: the system detects that the next part to be picked up does not correspond to the desired position and directs the robot to grip the part securely. When picking up sheet-metal parts from a conveyor, two-dimensional image processing is usually sufficient, since the height of the conveyor and the spatial position of the transport box are well defined. When parts have to be taken out of the transport box, an ISRA vision system working in three dimensions is required, since the position may also vary in depth, with two parts overlapping in one fixture.
Based on Proven Technology
Each function is implemented by a specialized system. Two-dimensional recognition is based on the Object Recognition Identification System (ORIS) technology, a robotic vision system that has proven itself in a range of applications, especially in the automotive industry. The three-dimensional system has 3D Location Measurement Sensor (LOCMES) at its core. LOCMES already controls a number of diverse assembly and joining processes in the automotive industry. Both technologies were developed by ISRA VISION SYSTEMS. The 3D image processing system uses a compact and robust sensor array. It has the ability to learn via a convenient Teach-in procedure which ensures a multi-functional versatility. Automated calibration processes use a reference object for machine setup. The user is guided via user-friendly window on the central Windows NT PC.
Savings on Multiple Levels
The potential for savings comes not only from lower downtime and drastic reduction in manual intervention, but also from an increase in cycle speed. This increase results from the fact that the robot does not approach the part at creep speed, but instead, grips the part quickly and securely as it is guided by the ISRA vision system. This also allows for lighter and simpler configuration of the robot arm.
Industrial image processing by ISRA VISION SYSTEMS lowers the number and duration of downtime and allows the use of cost-effective equipment in body assembly. This leads to short amortization times.
ISRA VISION SYSTEMS INC.
3350 Pine Tree Road
Lansing, MI 48911
our contact partner USA
Marketing Coordination: Diane Owens
Tel.: (+1) 517/887-8878 X 764
Direct: (+1) 517/887-3764
Fax.: (+1) 517/887-8444
For further information:
ISRA VISION SYSTEMS AG
Marketing
Industriestrasse 14
64297 Darmstadt
Germany
