Important Considerations for Your Next Shop-Floor Automated Inspectionby Joseph VanPelt on June 5, 2017 From Industry Focus
There’s no doubt that moving inspection out on to the shop-floor, closest to the point where the information is the most actionable, has huge benefits for any manufacturing process. Improved efficiency, better quality, reduced WIP (Work In Progress), and the ability to use SPC (statistical process control) are just a few of the reasons companies both large and small have been making this shift. While the design of most modern factories should help mitigate the impact of shifting to a near line or in line metrology approach, there are still several key concepts to consider to maximize repeatability, increase throughput, and minimize the error budget of the inspection.
The first, and perhaps the most important consideration, is the sensor technology used for the inspection. Usually the application dictates what types of sensors can be used. Start by reviewing your tolerances, and how much of the tolerance can be consumed by the error budget. Ten percent is the general rule of thumb, but sometimes up to 30% is allowed because of the cost to achieve the 10% allowance. Next, you should review your cycle time and environment, since the sensor technology impacts these functions.
Cycle time is sometimes a deciding factor in what type of system is used for an automated inspection cycle, especially if there is a requirement for 100% inspection instead of sampling. Some of the sensors in Hexagon Manufacturing Intelligence’s product line can bring huge benefits to cycle time. For instance, using a blue-light scanner or a laser line scanner can increase throughput if the application allows for the accuracy of these sensors. Other items that usually don’t get considered when looking at cycle time are load/unload and tool-change times. The load and unload time that a robot takes to place a part on a coordinate measuring machine (CMM), for example, is often not factored in when determining the cycle time of the cell. Likewise, the time it takes to change to different probes can also be forgotten when calculating the cell’s cycle time.
Environmental concerns can affect the amount of uncertainty budget that is used in an inspection system. Because of this, Hexagon’s shop-floor CMMs usually have two different accuracy specifications that allow for a different range of temperatures. Vibration, debris, oil, and other environmental factors can also impact the inspection process, and there are several ways that these items can be addressed. For example, the WLS 400 3D optical scanner that is used in the 360° Flexible Measurement Cell (360° FMC) can withstand certain amounts of vibration and temperature changes, making it possible to locate measurement near a hot stamping press. However, a CMM would require special isolators and a special enclosure to achieve the same proximity to large manufacturing equipment or heavy fork-lift traffic. One item that is often overlooked in regards to temperature is how quickly the temperature changes on the shop floor. A 0.1°C change every 10 minutes can cause problems with ultra-high accuracy CMMs, even if the temperature stays within 18°C to 22°C because the required accuracy would be affected by rapid expansion/contraction of materials. Some metrology equipment requires gradual temperature changes, and the sensor’s specification sheets should be reviewed carefully with a Hexagon representative if there are bay doors near the target area, for example.
Communication with a robot or production line is also a large consideration when planning an automated inspection routine. The type of communication usually dictates how the system is configured. Do we need to send Accept or Reject data to a material handler, for example? In this case, the system may make use of a PLC (programmable logic controller) to tell the robot to place the last inspected part on to the rework conveyor. We may also want to send feedback directly to CNC systems or other manufacturing equipment. One use would be to automatically adjust the CNC’s offsets in relation to the dimensions that are reviewed statistically with Q-DAS software. There may even be a threshold of change that signals the need for changing inserts on the CNC, which can remove the human error involved and prevent costly scrap or rework.
There are a number of subtle considerations an organization must review before embarking on an automated measurement process on the shop floor. Companies that are mindful of the specific topics that impact an automated manufacturing intelligence measurement system stand to gain the advantages of repeatability, minimized error budgets, and increased throughput – plus the strategic benefit of a fast ramp-up time for installation and utilization of the solution.
Joseph VanPelt works as a Systems Engineer in Hexagon Manufacturing Intelligence’s Integrated Solutions department. He has over 18 years of experience in manufacturing and automation, and over 8 years’ experience with metrology.