TestRun™

TestRun™ is used to confirm that a machine vision application is operating within expected parameters. TestRun performs tests on the In-Sight vision system and compares the results of those tests against expected values.

Note: TestRun is only available to users with Full Access privileges.

TestRun consists of the following tests for an In-Sight vision system:

  • Limit Tests: Looks at a list of parameters and determines if the current value is within the upper and lower limits, as configured for that parameter. These parameter settings can be configured by users through Controls functions within the job, discrete inputs read into the job or values sent by an external device through a communication protocol.
  • Image Tests: Sends a series of known quality images ("good" and "bad") to a vision system and determines if the results from the current job match the expected results for each image.
  • Hardware Tests: Compares the image quality of an acquired image from the vision system to the expected image quality to determine if changes in camera position, focus, lighting, etc., have occurred.

Why Test an In-Sight Vision System?

Machine vision applications are complex, with many operator-adjustable parameters, interrelated inspection tools, and environmental variables affecting the expected results. Testing the In-Sight vision system provides a “baseline” to insure that changes to one part of the system do not cause other parts to stop working. For example, the ambient light provided by windows during the day shift may affect expected results through overexposure.

In a production environment, TestRun provides plant managers assurance that the vision inspection system is catching defects, and not rejecting good parts. For vision system engineers, TestRun enables fast confirmation of the system’s performance when new tools are added. For maintenance technicians, TestRun enables parameter adjustments to be checked immediately, so they know if an adjustment in one area causes problems with other tools.

Set Up TestRun

TestRun assumes the following:

  • The machine vision application has already been developed and saved as a job file on an In-Sight vision system. A job file (*.JOB) is configured to perform a specific set of machine vision tasks. On In-Sight vision systems, job file storage is limited to the amount of available flash memory. File storage on an In-Sight emulator (running on a PC) is based upon the size of the hard disk in the PC. Job files can be loaded onto In-Sight vision systems from an In-Sight FTP server as well as from non In-Sight FTP servers. The name of the active job file is displayed in the title bar if it has been saved at least once. A job file can also be opened using drag-and-drop from the In-Sight Files pane or from Microsoft Windows Explorer.
  • Within In-Sight Explorer, a connection to the vision system has been established and the .job file loaded.
  • An image database containing the images used in the application has been created.
Note: Ideally, the image database will include examples of "good" images (images that result in a Pass) and "bad" images (images that result in a Fail). TestRun is able to operate with a variety of file organization structures for the image database, however, sorting the images into separate "good" and "bad" folders prior to launching TestRun will simplify the process.

EasyBuilder View and Spreadsheet View Developed Jobs

In-Sight Explorer provides two different programming environments for developing In-Sight machine vision applications (.job files): the Spreadsheet View and the EasyBuilder View. TestRun works with .job files developed in either environment. However, there are minor differences in the configuration of TestRun, depending upon which development environment was used. For example, TestRun relies on symbolic tags as references. In the EasyBuilder development environment, symbolic tags are created automatically as the job is built, whereas in the Spreadsheet environment, cell references are generally used (TestRun includes a feature to automatically create symbolic tags).