Sensors Integration Logo
Industry:
Food and Beverage
    Location:
    Delano, MN
      Customer Objectives:
      • Ensure that food package label matches the product inside 100% of the time
      • Execute 360-degree inspection of food containers on production lines 
      • Optimize throughput by minimizing product spacing on conveyors
      Key Results:
      • Reduced costs by about 30% through control design and using Cognex’s multi-reader sync
      • Using a smaller PLC with fewer IP addresses and ports instead of a more expensive model saved about $10,000
      • Reduced cost by more than $50,000
      • Reduced parts and complexity, including an improved supply chain, faster build times, and shorter integration periods
      Cognex Solution:

      Inspection in the food and beverage industry involves more than just making sure a UPC label matches the contents inside a package. Food inspection is about ensuring that what is on the package matches the product inside 100% of the time. Mismatches and other errors can potentially hurt a company’s brand, its finances — the average cost of a food recall to the producer is $10 million per incident — and even its customers.

      Years ago, a child of David Kotula — president of Sensors Integration, creators of innovative vision solutions to improve product quality — had an allergic reaction to peanuts. Fortunately, the family had an EpiPen and the child was unharmed. Driven by this near tragedy, Kotula is on a mission to perfect food and beverage safety. Among the results of those efforts is the Control Reliable Inspection System (CRIS). It uses four Cognex DataMan 474 barcode scanners with Edge Intelligence to provide 360-degree inspection of food containers on production lines, with minimal spacing while reducing PLC requirements and associated costs. In addition to optimizing throughput by minimizing product spacing on conveyors, CRIS has a unique “fail to safe” approach. It assumes that all product is mislabeled until it’s proven to be correct, resulting in unprecedented levels of accuracy in production and consumer safety.

      Assume the Worst, Achieve the Best

      Many inspection systems have set criteria for failed product detection. Such systems assume that every product is good unless one or more failures are found. However, according to Kotula, treating all products as good until failure is detected assumes that the inspection system is perfect. But many things can affect the accuracy of a vision system over time. For example, if a camera is slightly out of focus, it may not be able to detect failure criteria. A crossed wire at a PLC can lead to “failed” responses from an automated inspection system not coming through. More problems arise from failure to turn a downstream rejection system on after maintenance. CRIS addresses each of these concerns.

      Other problems involve the “teach” feature on many vision systems. Using the teach feature can accidentally teach the system to accept failed products. Additionally, without any logging access to the system, it is difficult to determine how long a line has been passing products that should have failed or who was responsible for making the false pass change.

      “One company had a line shutdown due to detecting failed product,” said Kotula. “The operator reset the machine, as they were trained, and it shut down again. Then a more seasoned operator suggested using a different button to keep the shutdown from recurring — which was the teach button. The company ran for two days with the incorrect image in the system, which could have been sending failed products out the door.”

      To prevent such errors, Sensors Integration omits the teach button. Instead, the company offers preloaded recipe selection screens only accessible from authorized personnel using their RFID card along with extra features for recording and time stamping events, which is imperative to figuring out why a product failed, validating that the system is running properly, and documenting any changes. Sensors Integration also added Cognex Edge Intelligence (EI) to the four DataMan readers in the CRIS system, enabling image archiving for post-event troubleshooting and system optimization and improving the customer’s overall process by connecting the system to a plant’s manufacturing execution system (MES) or enterprise resource planning (ERP) system. EI provides Sensors Integration an easy, fast way to connect to all their Cognex barcode readers for management and system optimization — a key design component of the four-camera CRIS system.

      In addition to image archiving, CRIS enhances operation accuracy by controlling access to the system through RFID badges with varying levels of access. Each interaction with the system is digitally recorded with a time stamp, personnel name, and more. Also, sensors are connected to reject equipment, letting CRIS know that all systems are online. After three consecutive failures, or in the case of any other wiring or actuator error codes, CRIS stops the production line and notifies the chain of command. Taken together, all these features have enabled CRIS to achieve accuracy greater than Six Sigma, in one test handling more than 6.2 million UPC codes without a single fail. The test was stopped at that point because the test containers were showing significant signs of wear from the conveyor!

      Food Safety for Less

      Safety is the primary concern at Sensors Integration, but automated systems must also fit within client budgets. CRIS is designed to be an easy-to-install retrofit over existing packaging lines — most of which were not designed with the spacing needed to accommodate automated UPC reading. The four cameras positioned around the packaging conveyor must be able to acquire images, find the UPC code, translate the code to a number, and pass that number to the controlling PLC — all within milliseconds. Cognex DataMan 474 barcode readers deliver high-speed reading capabilities with variable focus optics. These can adjust quickly to reduce or eliminate blurring and no-reads. However, to add all these safety features into a cost-effective package, Sensors Integration needed to go further.

      Multiple cameras increase complexity and costs and present a barrier to entry. However, with years of experiencing working as a Cognex integrator, Sensors Integration designs its control system to reduce complexity and cost wherever possible.

      Using a smaller PLC with fewer IP addresses and ports instead of a more expensive model saves about $10,000. Overall, Sensors Integration’s design reduces cost by more than $30,000. There are other benefits and savings in reducing parts and complexity, such as an improved supply chain, faster build times, and shorter integration periods. Using Cognex products and services, Sensors Integration offers a cost-effective design that consistently delivers Six Sigma quality or better. The general goal of food inspection is to keep people safe. CRIS is an advanced vision solution designed to verify product quality and validate product integrity. As the food and beverage industry increases production and speed, vision systems such as the ones offered by Sensors Integration and Cognex, are driving inspection solutions to keep consumers and companies safe.

      Sensors Integration has dropped costs by about 30% through control design and using the multi-reader sync feature from Cognex. This feature lets us use a single port, switch, and cable for an array of cameras, so we don’t have nearly the same PLC overheads as many other systems.

      ―President, Sensors Integration
      Sensors Integratin
      Featured Cognex Products

      GET ACCESS TO SUPPORT & TRAINING FOR PRODUCTS & MORE

      Join MyCognex