As profit margins shrink and consumer expectations rise, its more important than ever to optimize for quality in your operation, and reduce rejection rates. Fortunately, new smart factory technologies help manufacturers work smarter without increasing labour costs or materials costs. Today manufacturing organizations and their leaders can easily embrace available IoT-enabled smart manufacturing solutions
High-Volume CPG Manufacturer (Scenario 1)
The first manufacturer produces consumer goods products, using a mix of manual and automated production lines for mixing, melting and packaging. The production team is large and includes full-time and part-time machine operators, as well as shift managers, production managers, and process control engineers. They operate multiple plants. Each plant has multiple high-speed lines that produce about 100 units per minute. The average retail price per product is approximately $20.
Highly-Engineered Low-Volume Manufacturer (Scenario 2)
The second manufacturer produces pipes for oil pipelines, their manufacturing technology is based on metallurgy and coatings, their team is largely made up mostly of full-time staff, PhD’s in metallurgy and management team. The quality control processes typically take a few weeks to complete. The average price per unit, a pipe segment, is approximately $100,000.
Reject rate are a major problem in several scenarios:
Low salvage value
The difference between the retail price and salvage value (selling products in secondary markets) for ‘manufacturing rejects’ is 20% of the full retail value. Its a huge drop in revenue only magnified by the size of the production run.
Depending on the type of error, the rejected products can be fixed and sold as a final product. This can save materials costs, but, the labour cost per unit goes through the roof while disrupting other production orders. In many cases, it’s cheaper to sell them as ‘manufacturing rejects’ in secondary markets than to spend the extra time to fix it.
Finding errors late
On a high-speed production line, like the consumer goods manufacturer at 100 units per minute; when there is an error during production and no one catches it for 30 minutes; that is $60,000 worth of scrap that has to be thrown away.
On a slower production line, like the oil pipeline manufacturer at several weeks per unit; when there is an error during production and you find out at the end of the cycle, the impact is $100k for every bad unit in the run.
Production managers are searching for additional technology
to improve, beyond existing policies to ensure processes are followed, safety
specs are met, and sources of errors are minimized. Industry 4.0, Internet of
Things, Artificial Intelligence, Machine Learning, Predictive Analytics are popular
technology buzzwords, but they are looking for specific new IoT technology that
reduces rejection rates:
Production managers are searching for additional technology to improve, beyond existing policies to ensure processes are followed, safety specs are met, and sources of errors are minimized. Industry 4.0, Internet of Things, Artificial Intelligence, Machine Learning, Predictive Analytics are popular technology buzzwords, but they are looking for specific new IoT technology that reduces rejection rates:
Esprida LiveIntersect Solutions to Reduce Reject Rates
Sensors to monitor bins
This solution is quick-win; an easy way to add IoT sensors within your manufacturing process. The high-speed line at the first manufacturer had quality control sensors, however, they don’t have any intelligence around the waste bin that reject units fall into.
The problem is that the waste bin needs to be checked manually during the production run. Adding IoT sensors to the waste bins automates the manual process. Connecting them to LiveIntersect automatically detects the number and rate of error units. When there is a sharp rise in error count, then the LiveIntersect system automatically: turns on audible alarms; turns on a red light on the production floor; sends notifications to the shift manager via overhead screens and their smartphone.
Machine data solutions to enforce quality control
It is common to have variances between the official and actual standard operating procedure (“SOP”)/recipe for a production run. The official SOP specifies machine parameters for RPM speed, temperature, time for every production run; however, on a day-to-day basis, we’ve seen operators use this a guideline, they feel products will be better by tweaking these parameters because of changes in humidity, input materials and other factors.
These secret recipes are also a source of error, produce fluctuations in product quality, and increase
|Job||Operator||Mixing Speed||Mixing Temperature||Reject Rates|
|156||Jason||15 RPM||39C||5.7 %|
|157||Philippe||12 RPM||48C||8.0 %|
|158||Desmond||19 RPM||41 C||3.4 %|
Esprida LiveIntersect captures machine parameters directly from control panels and keeps track of production runs, records differences between results and parameters. Rather than asking the operators, yet again, to stick to official SOP after a failed run; you can show the data to operators that the production parameters are already optimized and shouldn’t be tinkered with.
Data from LiveIntersect can also be used to increase quality and/or profitability for the entire plant. If you haven’t already optimized machine parameters, then the data from LiveIntersect can be used to optimize for production time, production speed, product quality, product consistency. Implementing this as the new standard will improve plant performance without major changes to costs.
Predictive Algorithms to Reduce Reject Rates
Your existing production data can be used to predict failures. The challenge is often having the skill set and time to: cross-reference data from your multiple systems (ERP, Inventory, CNC machines, Milling Machines, Quality) and apply AI algorithms to find predictive patterns.
The predictions are based on analyzing operators, machines, jobs, and program parameters. Knowing the likelihood of failure helps you have a plan of action: operators that need guidance for certain jobs; products that are more sensitive to machine speed, or temperature
Real-time algorithms to reduce reject rates
In scenarios when machine data is available in real-time, you can prevent errors during a production run. Like predictive algorithms, real-time data from CNC machines, and other production machines, is used to calculate real-time predictions. These predictions can trigger warning notifications on smartphones or overhead screens to help shift managers and production managers make adjustments and reduce the likelihood of scrap.
Leaders in manufacturing can access the latest innovations in IoT sensors, machine data processing, artificial intelligence to increase margins, quality and more. This article provides a glimpse into just one opportunity (reducing reject rates) unlocked by the latest in smart manufacturing and IoT technology. Take your first steps in learning more about the potential and downloading this report. Use the abundance of data to unlock hidden insights.
This article was produced by Esprida Corporation. For more information on how smart IoT technology in your manufacturing facility, fill out this form.