The terms Industry 4.0, Big Data and Industrial Internet of Things (IIoT) are now on everyone’s lips. They describe the intelligent networking of extremely diverse areas in the industry. For example, in production & logistics, machines, storage systems and equipment exchange information with each other and can independently trigger actions.
But what does Industry 4.0 mean in practice for everyday applications?
It’s often the little things — ingrained habits that we hardly think about – that can make daily life difficult. Solutions are available today that would make a lot of these things more convenient. Things that are currently accepted as a matter of course in everyday life, such as e-tickets & mobile check-in using a smartphone or making an online purchase with a one-click order process, are convenient and can already be called standard. It is already standard for state-of-the-art cars to signal when service or an oil change is due. This depends, of course, on the driving style and usage behaviour rather than just the odometer reading. And so, this data is already automatically transmitted to the manufacturer or service shop. If this transparency may already go a little too far for some of us, industrial production is somewhat different. Big money and competitiveness are at stake. And such transparency certainly pays off.
Today, we still cannot estimate the ultimate extent of the ‘smart home’, which includes not only the controls for lighting, heat and blinds but also refrigerators that automatically reorder food. Though now it is clear that the consumer sector can’t imagine life without tablet, computers and smartphones, when Apple launched the iPad in 2010, just five years ago, many did not believe they could manage without a keyboard and operate it using just their fingers. Today, there are already more users swiping their fingers while on the go with mobile internet than users at desktop computers. We must pay a lot of attention to this to prevent industrial manufacturing from being left far behind by these trends.
The expandable modular concept
The moulds and tools used in manufacturing plastic parts are usually very complex and expensive. They are subject to wear & contamination and require regular care, cleaning and maintenance. Unfortunately, this care is often neglected due to lack of time and the associated effort & expense; likewise, the documentation of mould data, which often has to be entered by hand in mould sheets or books, is a time-consuming hassle. Moreover, these manual processes are also error-prone and lead to problems ranging from trouble with equipping all the way to mould breakage.
An initial goal is to reduce the effort for manual documentation or automate it, which is no longer a problem using RFID data carriers. Such data carriers, the ‘memory’ of the mould, can now be read using a commercially available smartphone with Near Field Communication (NFC) and, in turn, every mould into a ‘smart mould’. Automated mould management of injection moulds in manufacturing also eases preventive maintenance, reducing unexpected and costly production downtimes, making planning for the available resources more accurate. Usually, such solutions can be used by means of Manufacturing Execution Systems (MES) software, but only for newer, networked machines and often for a specific manufacturer. The option described in the following is meant to show a universal, non-proprietary solution that can be retrofitted without modifying the machine control system. This is not an isolated solution but a future-proof, expandable modular concept, with the outcome being a comprehensive lean production solution.
Manufacturing Execution System (MES)
The basis for state-of-the-art production planning and control systems is MES, which ensures, for example, that the machine facilities are working efficiently to capacity and the production quantities are manufactured according to schedule. Furthermore, manufacturing process data, which also contributes to quality assurance, is collected and managed. Such data is indispensable, for example, for regularly performed audits. Some machine manufacturers offer proprietary solutions, making sense if the machine facilities are uniform and consist predominantly of machines from one manufacturer. But if machines from different manufacturers are being used, including older machines, then non-proprietary solutions are better suited. An MES is a complex tool that provides extensive functions that are available to users but, in many cases, not at all to the full extent. An introduction requires good planning because it involves a significant investment, and though long-term cost savings are no doubt possible, the total savings potential is not easy to illustrate. Savings that can be achieved through MES, and thereby through lean production, consist of many different components. These primarily include lower throughput times & fewer rejects, more efficient utilisation of machines and optimisation of set-up time, to name a few. It is entirely possible to have transparent manufacturing in which IT systems, Operating Data Acquisition (ODA), material planning & order planning are networked with everything from the shop-floor management (MES) to logistics using Enterprise Resource Planning (ERP). Now, there are even standardised Key Performance Indicators (KPI), which are derived from the VDMA standard sheet 66412-1 and listed in the ISO 22400-2 standard. These include machine capability index, set-up ratio, utilisation efficiency, scrap ratio, technical efficiency, good quantity and Overall Equipment Effectiveness (OEE).
Machine Data Acquisition (MDA)
An easy first step in Industry 4.0 is the automatic collection of machines, production and mould data. A central metric for injection moulding is recording the shots produced. A large number of parameters can be determined using the shot count and the mould master data. This is the basis for the mould’s product memory. That takes care of the first step for efficient mould management. Solid, predictable maintenance depends on easy monitoring of service life, cleaning cycles and maintenance cycles. But here, the term ‘monitoring’ is already incorrect. Rather, the ‘smart’ mould independently tells the maintenance personnel when cleaning/maintenance is due. Production planning is simultaneously notified that, as of a certain time, the tool will not be available for an exactly defined period. This information is processed in production planning, and if needed, the usage of machines and moulds is adjusted, much like with a navigation device that calculates alternative routes when traffic is disrupted. Unlike preventive maintenance, the procedure returns reliable values, which can serve as a basis for predictive maintenance. Thus, it makes an important contribution not only to cost minimisation but also to comprehensive total maintenance.
Mould management – Life-cycle management
Although the inventory of injection moulds represents a large financial investment and usually far exceeds that of the machines, mould management that functions well is often the neglected child in an injection moulding business. Even companies that contract out their parts to contract manufacturers, making them mould owners, rarely have an accurate overview of their mould inventory’s respective status (asset tracking). Handwritten mould manuals, appended sheets, disks or USB flash drives are common, while databases and Excel tables are already the marks of a progressive shop.
The most expensive maintenance is usually the repairs after the damage has occurred with associated downtimes. Since many problems recur, one can use empirical values, provided they have been documented, and easily estimate when a certain error will reappear. The shot count is an essential factor for this. But the produced quantity is not all that is critical; blank shots when moving in also add to the wear. Likewise, the number of mould changes is important for maintaining the multi-couplings, for example. Many of these parameters can be stored directly at the mould to be called up everywhere at any time. This means the moulds are ‘smart’ and can save & provide this information.
Savings potential – ROI calculator
To enable a statement to be made about the ROI for such a solution, users were asked for their assessments. Some aspects from their practical experience were summarised, and the basic conditions were evaluated. Therefore, the ROI essentially depends on the number of available machines, moulds as well as the number of mould changes, the time for maintenance and documentation connected with this. Accordingly, the savings’ potential results from better capacity planning by the maintenance department and altered maintenance cycles. It is better to determine them based on actual production figures rather than on periods of time. In exemplary applications, this yields an ROI of less than one year. Of course, well-maintained moulds also achieve higher production figures than those for which maintenance is neglected.
Optimisation of set-up time
A major lever for cutting costs in injection moulding manufacturing is optimising and reducing set-up times. The more often moulds are changed, the greater the potential savings. Maintenance of the moulds themselves also often requires disassembly and causes several hours of downtime. Reducing these forced breaks or enabling them to be planned has a positive effect on the production process, too. The prerequisites for lean production are fulfilled by introducing an automatic mould change system. To ensure the use of the correct mould and corresponding gripper of the removal robot during an automatic mould change, these components are often already equipped with non-contact data carriers. Consequently, only a small step needs to be taken to use this technology even more effectively and simultaneously for mould management.
Cycle of an injection mould in production
Even without this degree of automation, however, automatic maintenance monitoring can be used successfully. This is shown by the following example of an injection moulding application in which the insertion of parts and the mould change are done manually.
Application example: In one manufacturing process, miniature sensors are over moulded with plastic housings. Two cassettes serve as moulds for this purpose; they are manually populated with the sensors to be encapsulated and alternately inserted into the main mould holder.
Problem: After some time, deposits pose the danger of the small, vulnerable fitting pins on the mould getting damaged or breaking while closing the mould.
This problem can be prevented by regular cleaning. Since cleaning during production leads to the mould cooling off, this would disrupt the production process and increase the scrap rate. Therefore, cleaning should take place only as often as necessary and, as much as possible, during scheduled breaks in production. This is why the mould was cleaned only once each time before the installation (mould change). But since this cycle depends on the produced lot size, cleaning may be done too infrequently (for large production lots) or too frequently. The lot size fluctuates between 1000 and more than 5000 pieces. In the case of large production lots, cleaning is usually not done in a timely manner. In the case of small lot sizes, cleaning is often even omitted. Experience has shown that a reasonable cleaning interval is about 3000 shots.
New solution approach
The shot count and various mould-specific process parameters are stored in non-volatile memory in an RFID data carrier due to the retrofit RFID solution. Detection of which mould is installed, additional mould master data and the current shot count are continuously counted and saved in the mould’s data carrier. A system with 13.56 MHz is used to enable easy mobile access via the NFC standard. Then the mould data can be read out at any time using a handheld scanner or mobile phone and can be called up online via web interface & web services. The data carrier can be easily fastened to the mould. Easy and individually adjustable options are available for installing the read head. An externally attached inductive or optical sensor, mounted at a suitable place on the machine, serves as the shot counter. The user can obtain the actual mechanical shot count by this means without accessing the machine control system.
After 3000 shots, the operator is shown a warning that changes colour incrementally up to 3500 shots to signal that the warning level is reached. Cleaning must be done no later than when the limit of 3500 shots is reached. Afterwards, a handheld scanner or Near Field Communication (NFC)-capable smartphone is used to reset the counter in the mould with password protection. A flashing red warning light will indicate if the resetting or cleaning was forgotten. Optionally, an e-mail can be sent to the production manager or machine setter in the event of a warning. All machines can be connected to the control level and an ERP or MES system using of LAN, Wi-Fi or Powerline. The result is direct access to the data and the process from everywhere.
With only the basic functions of this stand-alone solution, it is possible to access the application via a web server or connect to a mould database using JSON or MQTT. The range of possibilities also includes a professional mould management program or production control with standard MES software. This fulfils requirements for seamless automatic documentation, which customers are requiring evermore and also makes it easy to track parts or batches. One can move production into the IoT with a manageable amount of effort. This does not rule out further expansion stages; rather, one can use such a small solution to gain initial experience. It will then be easier for one to begin using MES tools as well as define and select further MES tools, helping to reduce the risk of investing in suboptimally dimensioned systems.