Vision Systems Keep an Eye on Critical Assets
| By: Winn Hardin, Contributing Editor
Wiki defines asset management as an accounting process that seeks to track fixed assets for the purposes of financial accounting, preventive maintenance, and theft deterrence. Not necessarily the first applications you think of when you build a machine vision system.
However, with the assistance of specialized software, vision systems can provide insight into the health of manufacturing equipment by analyzing statistical process control (SPC) data over time and analyzing it to reveal trends that reflect back on the manufacturing equipment itself. For instance, a product manufactured on a line with an automated machine tool will eventually suffer quality issues as the tool head wears over time. Therefore, monitoring the spatial variations in the finished product can tell manufacturers a lot about the health of their equipment.
Another connection between asset management and machine vision lies in the infrared (IR) spectrum. Infrared imaging has joined vibration analysis as a method to monitor the health of motors and other cost-intensive industrial assets, from beer kegs to shipping containers. By extracting radiometric temperature data from the IR images and automatically feeding this important data into computerized maintenance management systems (CMMS) used by maintenance departments around the world, machine vision can help extend the life of industrial machinery. Machine vision is also moving beyond the plant to help track critical assets by helping governments (and a growing class of specialized contractors) monitor the health of roads and railways – an application area that is accelerating thanks to government economic stimulus investments targeted at improving national infrastructures.
IR and CMMS
There are three main questions companies need answered by their asset management system: how many are there; where are they; and what is their status and value. Machine vision and imaging solutions, in general, assist with all three aspects of asset management.
‘‘We can take any photo associated with any piece of equipment regardless of format - still or video - and store it in our CMMS database for audit or ISO certification,’‘ explains Gamal Baladi, General Manager of MASS group, a company specializing in manufacturing automation software and systems.
While Baladi’s Traceability Made Easy (TME) CMMS system does not require images to function, many customers choose to include images of equipment in their databases for a variety of reasons. ‘‘Some people keep pictures or video to show maintenance procedures, calibration, or simply to show the condition of the asset. It may be 5 years in between services for a piece of equipment, so the more information you can give the technician through the database, the better off you are.’‘
More important to machine vision, or the process of extracting useful data from images, the TME system also accepts infrared (IR) images and radiometric data used in preventive maintenance programs. IR images show the heat generated by a piece of equipment. Electric motors, for example, will show hot spots as bearings and other moving parts wear out. The location of the hotspot can tell technicians where and when preventive maintenance procedures are necessary.
IR cameras from FLIR have been used to check beer kegs to see how much liquid they contain before cleaning and refilling, while engineers at Queensland University of Technology are one of several groups to explore optical character recognition (OCR) and bar code scanning to track these valuable containers, which can represent an investment on par with the brewery itself.
Sensors Unlimited, Inc., part of Goodrich ISR Systems (Princeton, New Jersey), a manufacturer of short wave infrared (SWIR) cameras for machine vision, has also sold their equipment to the wood pulp industry. ‘‘Pulp and paper companies use a furnace to reclaim the chemicals used to reduce wood to pulp,’‘ explains Doug Malchow, Manager, Commercial Business Development for Sensors Unlimited. ‘‘The recovery products build up a bed beneath the 2000-degree furnace and the plant needs to monitor that bed to make sure it’s building up evenly. Because of the temperature differential, they use a 1-m quartz optic. If the bed builds up unevenly, they need to do maintenance procedures earlier than anticipated to avoid problems. SWIR cameras are used because the lenses required by longer wavelength thermal imaging cameras can’t withstand the heat of the furnace.’‘
Protection is also a key part to asset management, whether it is through an offline maintenance program or during online production. For example, machine vision system provider PPT Vision (Minneapolis, Minnesota) uses its SENTINEL visible-imaging system to protect mold machines from damage by verifying that ejector pins are fully retracted before beginning a new cycle. If the pins are not retracted, the mold can be damaged, costing tens of thousands in repairs and many hours of downtime.
Vision Hits the Road
While manufacturers want to avoid downtime, machine vision systems typically avoid the outside where challenges in ambient lighting conditions can play devil with image processing algorithms. Road and railway inspection are two growing areas of asset management where vision loves the outdoors.
SICK, (Minneapolis, Minnesota) a provider of 3D and 2D machine vision systems, has sold its Ranger camera system to road and rail inspection contractors for many years, but business is looking up. ‘‘The government stimulus for infrastructure projects has dramatically increased the number of companies collecting road and rail surface data to sell to State governments for their federal infrastructure grant requests,’‘ explains Jim Anderson, Product Manager for SICK’s North American Machine Vision operations. ‘‘We’ve seen the number of customers doing this kind of work go from a just a few to 20 or more.’‘
SICK’s Ranger 3D camera is well suited to the task, being able to collect high-resolution 3D surface scans as well as 2D image data of potholes or other critical defects in a road or rail surface. ‘‘By using a combination of 3D and 2D data, these companies can use the 3D data as a flag and then go to the 2D data to see exactly what the patch, pothole, or rail surface is doing at that point. It saves a lot of time because these vehicles are collecting huge amounts of data as they drive roads or map a rail line,’‘ says Anderson.
SICK is also looking to use its Ranger technology to assist railway engineers by spotting and identifying switch signals. An engineer driving a train will pass three warning lights before each switch. A vision system could help warn a weary engineer of a switch, while also checking to make sure the light signals are working properly, essentially assisting the rail line owner to monitor his equipment.
While you won’t find asset management among the buzz words and marketing materials of most machine vision providers, when combined with data mapping, alarms, and mining software, machine vision technology regularly helps companies manage their most value equipment.