(2004), "Robotics and vision systems: the perfect combination", Industrial Robot, Vol. 31 No. 4. https://doi.org/10.1108/ir.2004.04931daf.002
Emerald Group Publishing Limited
Copyright © 2004, Emerald Group Publishing Limited
Robotics and vision systems: the perfect combination
Robotics and vision systems: the perfect combination
Keywords: Machine vision, Inspection, Assembly
Cognex's hardware and software solutions have allowed Thesys to develop robotic systems with integrated vision capability. These technologically advanced solutions are now in use in a wide variety of applications including object identification, quality control and assembly.
Thesys, an Italian company founded in 1989 with its main facility based in Brescia and headquarters in Montichiari is an expert in robotic solutions with integrated vision for solving specific problems in the area of object identification, quality control and assembly.
From the start, Thesys recognised that if it was to meet its customers requirements it would be necessary to give robotic systems a certain level of “intelligence”. In particular, this intelligence means an artificial vision system that is able to evaluate whether an assigned task is being completed correctly or not. With an integrated vision system, the robot can give an immediate feedback on the current process and supply valid and precise information about the correct way to carry out a particular task.
From closed systems ...
When Thesys started their business, there were already a number of vision systems available on the market, including those from Piaggio products, ABB and Kawasaki. “The major handicap for us” – explained Gino Storer, Head of Research and Development at Thesys – “was the fact that we are talking either about proprietary systems with limited functionality or systems which were not able to adequately support our clients' growing demands for greater performance”.
... to an open system
Because of the limitations of robotic vision systems available on the market at that time, in 1995 Thesys started looking for a partner who could supply vision solutions that would meet its needs.
The Cognex product portfolio also contains the most advanced vision tools available, which help guarantee successful solutions to all applications and technical problems.
“For example, for object identification applications, used to position the robot, Cognex had developed PatMax® and PatQuick™. These software solutions represented a real advance in the sector”. They radically changed robot guidance applications. PatMax can recognise and locate accurately parts despite objects and process modifications such as angle of view, scale, lighting and reflections. These vision tools use particularly powerful algorithms, which means that Thesys, can offer efficient and reliable solutions to all their customers problems.
Quality is controlled
A key business area for Thesys', to compliment their object identification and robot guidance activities, is quality control. In this sector, Thesys decided to enter into a relatively “virgin” market; metal surface inspection. As Storer explains, “We specialised in the analysis of components manufactured by both die-casting and sand- casting. After casting, these components are subjected to a series of work-processes including, stamping, sandblasting and washing. At the end of these processes, the components must have surfaces which are completely clean”. The checks and verifications to ensure this cleanliness are very complex, and high levels of precision must be guaranteed. These inspections include checking for the presence of stringers, dribble residues, deformations, oil marks and scratches.
In order to carry out these checks, Thesys has developed a system that is essentially a closed cabin (a cabin which is impermeable to outside light). The cabin is lit from the inside by a light system, which is perfectly stable and controlled. “In all the vision processes we use it is vital to have a high quality image, because the starting point for all of these analyses is to have a image that clearly shows any defect. The next important step, is to have a software tools which are able to interpret the image in the required manner,” said Storer.
From an operational point of view, the components enter these cabins on a belt conveyor. Then a robot, which is controlled by an appropriate programme, manipulates the objects in front of a series of cameras. These cameras are configured in the shape of a fan, so as to be able to view all surfaces within a limited cycle time. At the end of the analysis, the components, which have failed the tests, are separated from the objects that have passed. Those that have failed are classified on the basis of type of defect. From this information detailed reports are compiled.
The computer equipment used for these systems include an industrial PC, touch screen TFT display to provide the operator interface and a Cognex series MVS-8000 “frame grabber” based vision system. This family of acquisition cards is available in different models to suit particular applications, including both black and white and colour camera versions.
“One of the greatest advantages of using the feedback that the vision system provides” said Storer “is to help ensure the correct assembly of complex parts to form complete products”. A good example is the assembly of impulse transformers for telephone equipment. Owing to an innovative workstation design, made up of five robots and seven vision cameras, the assembly operations are organised in a highly flexible manner. From the monitor, the operator simply selects the product code to put into production. Each transformer is made up from a set of standard components including, ferrites, coils and plastic caps. In just 6 s, the entire product is assembled and subjected to a complete set of electrical tests to check 12 key operating parameters.
For further technical information, contact: Leigh Simpson, Cognex UK Ltd, Sunningdale House, 43 Caldecotte Lake Drive, Caldecotte, Milton Keynes, Buckinghamshire, MK7 8LF. Tel: +44 (0)1908 206033; Fax: +44 (0)1908 392463; E-mail: email@example.com