December 20208 MANUFACTURINGTECHNOLOGYINSIGHTSIN MY OPINIONQuestions of a Practical Vision EngineerBy Alex Reznichenko, Machine Vision Specialist, Edgewater Automation In the world of manufacturing, solution offerings are abundant. Yet for all of the effort in offering, most solutions rarely deliver the expected results. Machine Vision solutions walk a fine line along a very dangerous cliff of failure. For most facilities, machine vision is a "side job." Machine Vision solutions can quickly become complicated or complex. Hardware selection, software packages, and program methodologies each carries its own vast and rapidly evolving choices. Selecting and integrating the best solution for a specific application can become difficult and uncertain. For this reason, one must approach each application as a practical vision engineer. Practical vision engineers take time to understand the scope of the application and ask three fundamental questions that lead them to choose one of three tiers of machine vision solutions.Tier 1­Basic Sensor TechnologyThe first question a practical vision engineer asks is, "Does the application actually need a camera?" Yes, cameras can do presence/absence inspection, measurement, and checks for proper color but so can basic sensors. The advancement in sensor technology has made it possible for simple inspections to be robust and cost-effective. Very little reason exists to add smart camera if the inspection requirements are merely to check between two or even three colors. Sensors can do that for a fraction of the cost. Measurement applications can be solved by a laser sensor or simple photo-eye. In fact, I spoke with a customer who had an application that required certain work be performed on the product based on the measurement of a buss bar. The common solution discussed was to use a smart camera. However, after a quick discussion about the scope of the inspection, a much more cost effective and robust solution was discovered. Placing multiple photo-eyes in a row based on the different lengths of the buss bars and using the PLC to determine which length was present provided the same information with much less hardware and programming. There was no need for lighting or a large field of view or verifying of a vision program. It just worked. A practical vision engineer looks for reasons not to use a camera. An approach like this can produce a variety of successful out-of-the-box solutions. Tier 1 solutions look for the opportunity to utilize the ease-of-use and low-cost sensors for the given scope of an inspection.Tier 2­Simple Vision IntegrationThe second question a practical vision engineer asks is, "How many variations are there?" If the application is for a part with multiple components, and various features, and/or multiple colors then using a smart camera is a better choice. Smart cameras offer much more flexibility that can accommodate these types of variances. They key is to keep the inspection simple. Robust fixturing and strong tool selection are
< Page 7 | Page 9 >