This article covers:
This article provides a guide for factory owners and IT managers about the principles of lean manufacturing and the criteria to apply, in order to constantly work at optimizing factory outputs, and source the most cost-effective technology while reducing waste at the same time.
Many factory owners and manufacturers are faced with the challenge of transforming their factories from Industry 2.0 to Industry 4.0 smart factories in order to optimize operational efficiency and automation and to stay ahead in the competitive manufacturing space. Certain customers may require additional customization of products and faster output times, which factories also have to take into account. A large part of optimization involves leveraging and implementing new technology such as IoT architecture and Industry 4.0 systems while reducing waste. Implementing new technology in a factory can be quite an undertaking, and it is advisable for factory owners and manufacturers to avoid costly technology investments which yield no net benefit to the factory at hand.
During the previous decade, many factory owners moved to automation due to the benefits and gains such as higher accuracy, higher productivity, job scheduling ability and availability that increased mechanization offered. They often heard the phrase and the principle of “no islands of automation” that meant they were to avoid automated sub-systems that were not integrated into the overall factory processes and automation and thus provided no benefit to the larger systems in the factory. The aim was to have complete, integrated production and assembly lines that manufactured products seamlessly and without lag time. Automation in and of itself had a significant effect on the factory floor and factory owners experienced an increase in productivity and a decrease in downtime and lag time.
Now those same factory owners are hearing, “no island without a cloud”, since there is a push from IoT companies to promote cloud-based connectivity and solutions and store all the data the factory at hand is generating, in the cloud. The industrial sector is approaching standard cloud-based solutions with caution since there are concerns about the security of data, cost, bandwidth and latency. Even though the cloud does confer benefits to the Manufacturing Execution System (MES). Newer, emerging approaches are looking at using open standards such as OPC UA to control any machine in real-time and implementing machine to machine communication to reduce data storage requirements. The data is then collected and sent to a fog computer or processed at the edge closer to where the machines actually are located, to reduce the concerns with the standard cloud options such as cost and security.
Some of the key functionality related to Industry 4.0 technology that suppliers can provide and factory owners should take into consideration are:
Data-driven plant performance optimization refers to collecting and using data generated by the factory machinery, sensors, HMIs, PLCs, staff and SCADA systems in order to enhance plant operations and processes. The data cycle for plant optimization involves recording and monitoring data, uploading data, analysis of the uploaded data and the reporting of this data using IoT gateways and IoT architecture in the Industry 4.0 context. This optimization should strive to maintain Overall Equipment Effectiveness (OEE), which is a measure of how effective the plant and its industrial equipment are. A process that receives a 100% OEE score means that it has a high-quality output that is as efficient as possible with no machine downtime.
Data-Driven Inventory Optimization refers to the process of using real-time data to manage inventory. For example, consider a construction industry scenario where units of supply are labelled with RFID tags and an IoT system can count them. As soon as the supply units drop below a certain level, the sensors trigger an alarm and more supply units are purchased. Consequently, downtime is avoided and the project is more likely to be completed in the scheduled time frame.
Due to the ability of IoT systems to collect and manage big data, the IT provider should provide software that is able to develop quality-control models and profiles based on the data. Therefore, each product can be compared in real-time to these profiles (which were based on thousands or hundreds of thousands of data samples) and either rejected or accepted.
This model allows factory owners to turn their machines into stand-alone income generating streams, in addition to the revenue the machine generates from being part of the internal factory processes and production line. So in this model, a specific machine in the factory can be outsourced to a customer or another company that needs it for a set amount of time, and this customer can, through the IoT platform, receive real-time data about the products or services for which they are using that particular machine. A technology supplier should be able to provide HMIs or other systems that enable this multifunctionality. So the factory should be able to receive data about the internal processes the machine is part of and the company hiring the machine should also be able to receive data about the machine and its outputs relevant to their needs.
The Human Data Interface refers to the platform used for humans to engage with the data, this could be via calls to a database, an HMI, or even a smartphone. The technology provider at hand should provide an interface that allows personnel to engage with the data and draw insights from it.
Predictive maintenance refers to the use of data generated by a certain machine, in order to predict the chances of failure of that specific machine before the actual failure takes place. The maintenance of the machine then takes place proactively rather than reactively. This reduces downtime significantly.
Remote service refers to the ability to remotely monitor or repair machinery. This allows repair and maintenance to take place from anywhere and saves the factory owner the cost of transporting machinery to a repair site to be fixed.
Virtual training refers to training that is provided in a virtual capacity through the use of AI glasses. So, personnel can access this training and learn more about the factory processes in an online environment. Validation refers to the ability of the IoT system to check that the training received was actually beneficial to the staff and the factory. This is done by using sensors to compare the finished products of the factory before and after the completion of training, in order to see if there is a positive difference. Validation also involves using AI glasses to see if the staff member is actually implementing the training received on the shop floor.
Lean manufacturing is based on the concept of eliminating waste from factory processes while ensuring that the customer or client receives the maximum value. Lean manufacturing looks at optimizing the delivery of products in horizontal value streams that ultimately connect to customers. It is about evaluating what is adding value to the customer versus what is adding waste or is not beneficial to the factory.
It is systematic and there are five main principles involved in lean manufacturing:
Lean manufacturing principles can be beneficial for factory owners since they can be used as a technology filter or criteria in order to ensure that any technology implemented in the factory contributes to the reduction of waste and horizontal value streams. The technology in other words should contribute to the reduction of waste, the reduction in standing inventory, increased factory outputs, decreased production costs, and increased labour productivity.
Data-Driven Plant Performance as discussed above refers to the use of data in real-time to increase production. This happens simultaneously while using the data to identify areas of waste and unproductivity. Data-driven plant performance contributes significantly to all the five main lean manufacturing principles since customers receive value, the mapping of the value chains are guided by actual data received in real-time, and the data helps identify the barriers such as when there is downtime and which machine/process is causing the downtime, so this can be instantly rectified. Additionally, since there is constant delivery of data from multiple sources in the factory to the staff and personnel of the factory – they can develop pull systems due to the ease of communication and the constant analytical processing of the data. Furthermore, the continuous development of useful models based on big data and real-time data allows for continuous improvement.
Data-driven quality control as mentioned above looks at comparing a sample or material to a profile developed from big data rather than conducting many expensive quality-control tests on every single sample in the production line. This fits in with the concept of lean manufacturing since the number of tests is reduced but quality control is maintained.
Virtual training and validation look at providing training in virtual environments using AI glasses and validating through the use of AI glasses that the training was beneficial, effective and actually implemented. One of the main aspects of lean manufacturing focuses on training staff about lean principles in the factory since staff are a critical component in any factory environment. Therefore, through the use of AI glasses, staff can be trained and guided on lean manufacturing principles in the factory environment they are operating in. Additionally, the AI glasses can validate that staff actually are implementing the training they received in the factory. Consequently, the lean manufacturing concepts of waste reduction and optimization of product delivery will be felt throughout the factory as a result of both virtual training and validation.
Industry 4.0 concepts such as connecting multiple machines, machine-to-machine communication, human-machine communication, real-time data delivery, big data processing and analytical operations really tie in with the fourth principle of lean manufacturing.
Most manufacturers not using lean manufacturing principles rely on a push system which is based on standard forecasting techniques. Production is aligned to those pre-determined set forecasts. This can be problematic since some standard forecasting techniques are inaccurate, increase waste and are not effective. The lean manufacturing pull principle of not producing anything until there is a demand relies heavily on effective communication. With the correct choice of Industry 4.0 technology, this effective communication system can be developed and thus reduce waste and optimize overall factory efficiency.