Product category:
Vision and Colour Sensors
News Release from: DVT Sensors | Subject: SmartImage sensors
Edited by the Engineeringtalk Editorial
Team on 16 January 2004
Sensors maintain quality of latest
"hogs"
As Harley-Davidson and Rockwell Automation celebrate 100th anniversaries, SmartImage sensors start work inspecting the popular V-Rod motorcycle at Harley-Davidson in Kansas City, USA.
As Harley-Davidson and Rockwell Automation celebrate 100th anniversaries, DVT SmartImage sensors start work inspecting the popular V-Rod motorcycle at Harley-Davidson's Vehicle and Powertrain Operations in Kansas City, USA The V-Rod features the impressive 115hp, 1130cm3 Revolution engine, developed by Harley Davidson in a joint venture with Porsche
This article was originally published on Engineeringtalk on 16 Jan 2004 at 8.00am (UK)
Related stories
Smart cameras for smart engineers
Plant engineers face a variety of challenges: improving quality; increasing efficiency; managing costs; reducing scrap; reducing downtime; and boosting capacity.
Image sensors sove a sticky problem
SmartImage cameras have been used to eliminate expensive waste inherent in the production of corrugated cardboard cartons, while simultaneously fine-tuning one manufacturer's production line.
Some of the V-Rods are limited editions that include special 100th anniversary emblems.
All V-Rod assembly is performed at the 30,000m2 Kansas City plant, which employs some 930 people on four lines.
Two of the lines are dedicated to V-Rod assembly: one for engine assembly and the other to complete production of the unit.
All four of the lines use SLC 5/03 Allen-Bradley PLCs as controllers.
The inspection, which has been running since February, looks at bearing and balance shaft presence and proper snap ring seating in the engine transmission block.
Further reading
Easy Ethernet connection for vision systems
DVT Sensors launched three new products at Total which solve any potential connectivity problem relating to smart sensor installation.
Barcode software handles lower light levels
A move toward gradient pixel base threshold in FrameWork 2.7 has enabled more intelligence in the software to compensate for low-light levels and for variances of light in specific applications.
True robo-vision arrives
DVT Corp and Kuka Robot have developed a robot with seamlessly integrated machine vision.
The Allen-Bradley PLC has been used as a controller for the engine assembly line since it began producing V-Rod engines in 2001.
DVT Corporation's 530 SmartImage Camera, tied in to the Allen-Bradley PLC, verifies the high quality that is required.
When system integrator Chris Davison of Bachelor Controls, an authorised Rockwell Solutions Provider, got a call from Terry Carson of the DVT distributor company Power Motion Sales, he knew this application would be more challenging than the existing snap ring inspection in place on the V-Rod line.
For that inspection, the engine part was positioned on the line in a manner which allowed easy access for the DVT camera.
The new inspection, however, involved an engine block in a different position on the line.
The working area was too tight for the small 112 x 60 x 30mm form size of the camera.
"When I first learned of this project, I knew it would require more creativity and resourcefulness than a standard application", Davison said.
"This kind of challenge is exciting to me because it can potentially open doors for other applications".
Bachelor Controls worked with Harley-Davidson Manufacturing Engineer Dan Bruyn to come up with an innovative solution.
The camera was attached to a linear actuator which is servo-controlled.
The actuator is then connected to a pneumatic slide with the whole unit able to extend to three different positions to inspect three identical snap rings in different locations on the transmission block.
The inspection procedure involves the camera extending from the home position to the first inspection point, moving back to the home position, and then back to inspection point number two and then point three, each time returning to the home position.
This inspection procedure is performed simultaneously with the assembly of 20 internal engine parts that occurs at that place on the line.
(A required criterion for the application was that no time would be lost on the line).
As the camera could not get close enough to the inspection area, a mirror was affixed at a static position on the pneumatic slide between the press and the engine block to allow the camera to capture images.
The position of the mirror had to be situated at an exact angle for the camera to take precise measurements.
Once the pneumatic slide, camera and mirror assembly were complete, another challenge presented itself: the bearing and balance shaft inspections and one of the snap ring inspections were routine, but the position of the other two snap rings created lighting issues.
The camera could not verify seating of the snap rings.
After extensive use of the DVT Framework "blob" tool and associated software filters, Bachelor Controls and Bruyn devised an inspection process with total repeatability that sufficiently addressed all lighting and image issues.
The use of the software filters compensated for the lighting conditions and other issues to a degree where the images captured were high enough quality to reliably provide all needed data.
Another challenge for the application was identifying a location to be used as a reference point for camera measurements.
The snap ring is in the shape of an incomplete circle, similar to a round horseshoe, with two holes in the ring near the part where the gap begins.
The first thought was to use the two holes to establish a fixed location.
Instead, Bruyn and Bachelor Controls decided to use the gap in the ring as a reference point, measuring two edges on either side of the ring where the gap begins.
After the gap was located, a heavy amount of scripting was required to move the location of multiple line gauges.
This involved transferring a spherical coordinate system to a Cartesian co-ordinate system to offset the lines to ensure they would hit the snap ring edges in the correct locations.
Using the DVT Emulator tool, which permits users to work on applications offline, the two engineers viewed hundreds of logged images of the application without having to wait for images to come up in the actual process.
"The emulator tool helped immensely because it meant we didn't have to babysit the line as it cycled through.
This sped up the process dramatically", Davison said.
Once the machine vision unit was set up, Bachelor Controls added a remote I/O module to the Allen-Bradley PLC to communicate to a new remote I/O flex rack.
It added two power supplies - one for the Flex I/O and the other for the DVT equipment.
All digital I/Os were connected to the DVT breakout board.
The DVT breakout board I/O was connected to the new Flex I/O for station sequencing.
"This PLC expansion was a straightforward integration with no major challenges and it has worked flawlessly", Bruyn said.
"Bachelor Controls built a new panel with the DVT I/O board connected to new Flex I/O and provided the programming for the linear actuator, pneumatic slide, DVT communications and miscellaneous I/O through this Flex I/O", Bruyn noted.
DVT invented the smart camera in 1990, introduced Ethernet capability in 1998, produced the first megapixel high-resolution camera in 2001, and produced the SpectroCam in 2002 - the first smart camera to use a prismatic system for extremely precise colour measurements.
DVT is also the only machine vision company that offers free training, free technical support, and free software updates.
Small enough to fit in the palm of your hand, DVT SmartImage sensors are simply nodes on a network that acquire and process data to robotics and motion controllers, report inspection results for QA and deliver part tracking and inventory control.
Using standard Ethernet communications, SmartImage sensors support TCP/IP, Modbus/TCP and EtherNet/IP.
Profibus and DeviceNet are supported with the optional SmartLink module.
These self-contained inspection systems provide a visual data source with low connectivity costs, high speeds, seamless data exchange throughout the smart factory enterprise and enhanced tools to support automation locally and worldwide.
• DVT Sensors: contact details and other news
• Email this article to a colleague
• Register for the free Engineeringtalk email newsletter
• Engineeringtalk Home Page
Search the Pro-Talk network of sites
