Tools You Can Use Today to Build the Plastics Processing Plant of TomorrowTools You Can Use Today to Build the Plastics Processing Plant of Tomorrow
Forward-thinking plastics processors are applying artificial intelligence, advanced sensors, next-gen robotics, and data-driven decision making to create new thresholds of efficiency.
December 10, 2022
Advanced technologies such as artificial intelligence (AI), Big Data, and high-tech sensors are changing the dynamics of plastics processing by providing tools that learn from one another and make decisions to improve processes.
Plastics processing equipment with automated technology enables the equipment to communicate and process data with little or no human intervention. These embedded systems are integrating operating technology and information technology to monitor and control physical processes.
A joint study by the Manufacturer’s Alliance for Productivity and Innovation and Deloitte published in September 2019 found that more than 85% of industrial manufacturers believe that smart factory initiatives using advanced technologies will be the main driver of manufacturing competitiveness in the next five years. IDTechEx Research estimates that companies will spend more than $250 million by 2032 for industrial applications of printed and flexible sensors.
Sensing material, process behavior
One example of advanced sensor applications in plastics processing is SensXPERT technology from Germany’s Netzsch Process Intelligence. The technology analyzes material behavior with in-mold sensors that allow dynamic, adaptive production by reacting to material deviations, the company said. Customers can connect sensXPERT to existing manufacturing and control systems with standard interfaces, or use it as a cloud-based service.
SensXPERT technology analyzes material behavior with in-mold sensors that allow dynamic, adaptive production.
“The sensXPERT package is designed to deal with every machine using modern industrial interfaces like OPC-UA, PROFIBUS, PROFINET and — in the case of retrofitting older machines — analog and digital input/outputs,” SensXPERT Managing Director and CTO Dr. Alexander Chaloupka told PlasticsToday. “The hardware connection is usually the easiest process and the end-to-end communication has to be set within the software to allow sensXPERT and machines to talk to each other.”
The system includes several pieces of hardware, including two dielectric sensors installed in the mold and an Edge Device, which is hardware outside the press that collects data measured by the sensors inside the mold and process parameters from the press itself.
Simulate, predict, analyze
The Edge Device assesses hardware and software to produce models that capture minute deviations in materials and process. The resulting algorithms simulate, predict, and analyze material behavior on individual machines. The algorithms’ ability to parse huge amounts of data flowing in from molds in real time helps the system get smarter over time. If the algorithms flag a set of data as different, machine learning software alerts the technician monitoring the press that something is changing. This allows the operator to decide if action is needed.
Key parameters like glass transition temperature, pressure, and curing requirements “train” these process models, which are continually refined. This data-driven model is designed to improve the quality and efficiency of manufacturing processes.
SensXPERT said its technology will work with a wide range of materials — including thermosets, thermoplastics, and elastomers — and methods — injection, compression, and transfer molding; thermoforming; vacuum infusion; and autoclave curing. A web app allows users to access the system remotely.
“We have historically grown our business with thermosets,” Chaloupka told PlasticsToday. “Due to the complex chemical reaction, there is a high demand for process transparency and automated control to avoid scrap production and to work at the limits of the process. Within thermoplastic processing, similar to what we observe during the solidification of thermosets, we can see the crystallization and temperature behavior in the mold and use third-party sensors to measure the in-mold pressure, as well.”
AI lets operators see into the future
Trying to identify a plastics processing problem manually requires a focused team effort pulling data from many different sources, including oil analysis, vibration analysis, sensor data, and electrical testing. That data has to be analyzed, which is a reactive exercise that is labor intensive, tedious, and time consuming.
Harnessing AI to gather and analyze the data in a multivariant analysis can reduce the time it takes to discover problems, according to Dominic Gallello, CEO of SymphonyAI Industrial. Instead of reacting to a problem, operators can examine the data in real time, which is more beneficial than looking at data after the fact. AI’s capacity to assess a variety of factors and produce a comprehensive view of operations and processes also gives operators the vision they need to see into the future, he said.
A key aspect of the factory of the future is the increased use of sensors that can measure every aspect of plastics processing, including temperature, vibration, speed, duration, pressure, proximity, smoke, and humidity. “But data isn’t enough,” Gallello wrote in a recent PlasticsToday article. “It takes AI to gather, analyze, match, and assess the data to produce meaningful and actionable insights. Together, AI and wireless sensors can gather and investigate enough data over days, weeks, and months to accurately predict hard-to-detect gear wear, early bearing wear, and other critical faults well before they happen. They can make precise forecasts using multi-dimensional models that far exceed what engineers can do with univariate models. And as advances in computing technology, sophisticated AI, and machine learning provide more accurate results with consistent data, they also greatly reduce the burden on people to act as detectives.”
Processors only apply about 2% of available data
Plastics processors aren’t exploiting approximately 98% of the data available to them about their operations, Gallello contends. By using sensors, AI, physics, and embedded domain expertise, they can better understand their operations and processes.
While standard in-line sensors are built into most plastic manufacturing machines today, they are not always adequate to achieve the highest levels of benefit from process optimization, according to Prashant Srinivasan, Director of AI Products at SymphonyAI Industrial.
“As melt pressures exceed 150 MPa and temperatures are frequently above 300°C, in-mold pressure sensing elements are exposed to harsh conditions in a corrosive and abrasive medium,” Srinivasan told PlasticsToday. “A number of new technologies, like wireless thin-film piezoelectric sensors, are now available in the market. For temperature sensing, standard shielded thermocouples are subject to significant lag, making it desirable to consider installation of IR-based temperature sensors. Likewise, for online quality measurements, advanced AI-based automated vision systems are now available and may be installed to achieve the highest quality control capability.”
No-code robot programming
Sepro Group, a leader in injection molding machine and robot integration, is now working on “no-code” programming, with a robot controller using AI to optimize trajectories and manage obstacles.
At K 2022 in Germany, Sepro allowed booth visitors to reposition a simulated mold and other peripherals and then challenge Sepro’s AI solution to calculate the best possible trajectory depending on which of three primary objectives — maximum energy savings, minimum wear, or fastest cycle time — is selected. The system calculates ideal trajectories before the cycle starts, without any operator-written code.
Interactivity was a key focus for Sepro at K 2022 in Düsseldorf, Germany, in October. Shown here is a demonstration cell that gave visitors a chance to choose the best human machine interface to pilot a Sepro S5-15 Speed robot through a series of motions.
The energy-saving mode can reduce energy consumption by as much as 25% on certain trajectories, which is ideal for processors seeking to reduce their carbon footprint. The minimum-wear mode reduces stress on system components, which prolongs their lifespan and reduces maintenance.
Also at K 2022, Sepro’s new centralized control software, Visual+, was at the center of a multi-step production process featuring a 110-ton Milacron injection molding machine and two Sepro robots. The process included assembly of toy sailboat components, ink-jet printing, production-data collection, dimensional check, tray packaging, and delivery of the finished boats to booth visitors using an autonomous mobile robot.
The new controller uses an open communication system that allows better synchronization of complex movements, integrated peripherals, data management, and traceability. It also can communicate seamlessly with almost any brand of molding machine or secondary unit, the company said.
As an open system, it links to the controls on molding machines and peripheral equipment using a single centralized and intuitive human-machine interface for more intuitive machine operation and an enhanced user experience, the company said. It can collect vast amounts of data from all connected systems, which can be used for process optimization, traceability, and analyses for calculating overall equipment effectiveness and other metrics, locally or in the cloud.
Don't be left behind
Using sensors, AI, multivariant analysis, and machine learning, manufacturers can harness the power of vast amounts of data to build models to predict product quality as a function of process conditions and settings. These models can be used to optimize and recommend settings for a given product to achieve the optimal quality and avoid rejects. They also can automatically learn from new data and adapt to aging machines and changes in operating conditions. Manufacturers that can effectively use these high-tech tools will reach new levels of efficiency.
About the Author(s)
You May Also Like