Lean initiatives yield significant improvements in factories around the world by removing redundancy and waste in factory layouts, ergonomics, workflows, processes, procedures, and assets. These are generally tangible and visible elements of the factory and its processes. Today, many manufacturers are applying the philosophy of Lean to their supporting information systems. They are taking an objective view of the transactions and overhead involved in preparing data for the factory, launching products, gathering data from the process, and reporting upon it. These enterprises are finding that while the R&D and ERP systems are nicely consolidated and integrated, when they turn to the factory, they sometimes find up to hundreds of different software tools, systems, and databases each trying to support the overall scope of operations.More concerning is the common discovery that the engineering and IT personnel required to simply generate reports, collate data, and maintain the disparate systems is enormous and excessive when compared to the value and timeliness of the reports generated. Once the traditional Lean opportunities in the factory are addressed, there typically exists great potential for Lean improvements in the often invisible flow of information throughout a factory.
The enemy of a Lean information environment is redundancy. Multiple systems into which similar or even identical information are input in order for them to perform their individual functions, and then to consolidate the results from these systems, even more redundant labor is required to merge the information and make it visible. Consider a typical scenario; CAD and BOM data comes from the customer. It is used to create visual aids; to generate printer, placement, AOI, AXI, and test programs; to drive whatever setup control systems that exist around the factory; and to set up the quality collection systems. Every element requires the same data, and yet in many factories these systems all operate independently and the data is input to them redundantly.
Now consider the resultant data once the process executes. The information is now compiling in multiple databases. The machinery is logging events. Operators are inputting data to the quality system. Test and AOI are gathering similar data but in different databases. The operational overhead of all these data input tasks and data output compilation tasks is enormous. Furthermore, the maintenance of all these disparate systems is error prone and costly. Fortunately, these costs can be reduced significantly through system consolidation. The enterprise benefits from greater speed, process control, and data visibility as a result.
Today, as business leaders in manufacturing enterprises look to do more with less as well as increasing quality and visibility to the operations that actually produce their value, their Lean efforts are looking toward manufacturing information technology. This has led to a new category of “enterprise level” software called manufacturing operations system (MOS). To explore the purpose and value of a MOS system is to consider first how it naturally emerged – as result of market needs and conditions. The value of consolidating software solutions was appreciated long ago on the manufacturing enterprise’s business side. This drove the scope of ERP into virtually every aspect of business and financial management operations. The past decade has seen massive investments in such systems by business leaders in all forms of manufacturing enterprises. Today, these companies would likely shudder at the thought of not having a single integrated solution for their business operations.
The next evolution involved R&D. Once the financial and business operations were running more efficiently as result of ERP investments, business leaders turned their attention to time-to-market. Taking a product concept to a finished design affected competitiveness. Product lifecycle management (PLM) systems brought design, collaboration, version control and simulation under one massive software system. Millions of dollars were invested in these systems as well.
Then came the final problem – manufacturing enterprises, having invested hundreds of millions of dollars in ERP and PLM, began to consider why they continued to have key business issues. The business offices ran well. Financial reporting, customer management, accounting, sourcing, and even HR were improved. R&D was turning out better designs faster and under better control. Information access was greatly improved. But despite all this, the enterprise was still too slow in ramping to a quality, full-speed production run. Changing designs mid-production or even between products or jobs was too slow. And once products were shipped, traceability and quality metrics were cumbersome and slow to obtain.
In summary, companies’ time-to-market was fast due to PLM and ERP investments, but they realized that to ship the product, the factory had to ramp and run efficiently. The problem, it seems, was lengthy “time-to-value” – the time required to get a quality product shipped to a customer so it can be turned into revenue. Finally, management’s attention turned to factory operations.
The problem was rooted in disparate software tool proliferation. This occurred because rather than addressing factory information in a holistic manner as PLM and ERP addresses R&D and business, operations investment in software had been reactionary. For each requirement placed on the factory, a new software system was spawned. When traceability was demanded, a line-level software tool was purchased. When process planning and launch was too slow, a CAM tool was purchased. The test department logged its data in a system they purchased, while the quality department ran a data collection tool of its own design. This story has many variants, but the theme is always the same. In many cases, manufacturing information systems do not operate on one data source, and there is no means to monitor, view or mine the information across all of them.
In reaction, IT departments or integrators have tried to knit these systems together. This has led to significant staff overhead dedicated to designing reports and other tools intended to collate this disparate data. This means internal IT engineering or third-party integrators – and more overhead and higher internal maintenance costs.
THE THIRD PILLAR
There is now a desire to have what might be called the “Third IT Pillar” in the enterprise – the software system devoted to manufacturing. This concept has come to be called MOS. Its goal is to exist as a complementary system to ERP and PLM and to consolidate the systems and visibility to everything that goes on within the factory office and factory floor.
Consider first the financial benefits of consolidation of these disparate manufacturing systems. When a single system covers manufacturing operations and, by nature, provides seamless reporting, monitoring and analytical data, the enterprise needs only to build an integrating bridge among PLM, ERP and this system. Rather than hundreds of integrations and databases, the task of integrating only two major systems is manageable, and consequently much less expensive to build and maintain.
Exploring further into cost savings, the enterprise need only form a relationship and invest in maintenance with one vendor rather than many vendors. In terms of data fulfillment, the legions of report writers and IT personnel typically employed to develop data for engineering, regulatory agencies, customers or executives can turn their attention to other value-added tasks, as the information is now available through one system.
The operational benefits of MOS are vast. When one considers manufacturing software in the context of “Lean,” there is no more efficient way to handle information than through one system and one data source. MOS spans so many functions in the operation that its very operation is by nature faster and easier. There is no file passing or transfers, or data migration from one tool to another. There is no need to join data to achieve one view of information. Everything flows from the design department through a process planning engine to the floor to control materials and production movement, acquires real-time data from the machines and test systems, and then presents the byproduct of all these processes in real time or analytically. Indeed, the concept mirrors the essence of Lean.
Electronics assembly presents unique challenges compared to processes such as box-build or system integration. Planning, documentation, machine programming, shop floor materials control and even bills-of-material management require more depth in electronics than seen in more generalized steps across a box-build process. MOS in electronics begins with a rich awareness of the corporate part master, inventory master, customer part numbering relationships, BoM versioning, and design versioning. The baseline information sources that descend from business and R&D systems must be reflected in the core data model and systems of the MOS solution to even begin its operation.
Planning
The first phase of operation involves planning. When CAD design data and BoM for a particular work order are applied to process awareness that spans an entire factory (or multiple factories), the MOS system assists process engineering in creating a total plan for the manufacture of that product. This planning phase of MOS includes the design of the route flow, the component and activity allocation across the route, the “bill of process” for each step against which materials will be verified on the shop floor, operator visuals, machine optimization and programs, quality and test operations, and supporting reference materials to be presented to operators, when required. This becomes the assisted development of what is essentially a “digital work package,” providing the factory every conceivable guidance, verification and information needed to conduct a quality build efficiently.
Controlled shop floor execution
Execution on the shop floor is controlled and monitored. MOS enables control of both axes of production, the product flow through the route, and the components, consumables and tooling that must intercept that flow at the exact right time and location. Components, feeders, carts, tooling and chemicals are tracked and verified, ensuring a correct build. The flow sequence and activities along the route are also enforced, but the system also dispatches interactive operator visuals to each station automatically, or deviation and ECO notifications automatically, providing a fully paperless environment. MOS drives quality data collection, and repair automation and diagnostics by providing operators the visuals and context data they need to manage in-process quality efficiently.
A data-rich environment
Underpinning the visible activities of the process is the equally important element of MOS, comprising of the real-time data acquisition from machines, test, and other data sources. The database on which the solution rests, however, needs more than the product and materials movement information. It requires real-time data from assembly, process, test, monitoring and inspection systems. Most important, the solution must provide a native ability to conduct materials verification, WIP tracking, etc., and also be able to leverage existing machine systems sometimes provided from the vendors, such as feeder setup solutions. Rather than replacing such tools, MOS can add value by folding their functions into the overall solution as a data source, thus leveraging rather than disposing of prior investments – but yielding the same functional and informational result via MOS.
Analysis, reporting, monitoring, traceability
These should be the natural byproducts of a solution involved in the aforementioned functions. MOS approaches the issue from the perspective that a totally controlled and monitored process will naturally yield any type of traceability desired. To make such a system valuable, it needs to empower its users to view, analyze, and report any data they require without having to consult their software vendor or IT department. The information also should be available via real-time dashboards or through data extraction and reporting systems that do not expect the user to know anything about databases, SQL, etc.
The full business value of such a holistic system is realized when the factory is running real-time dashboards, and engineers and production planners are analyzing their quality and production information completely on their own. In the best deployments, plasmas above factory lines display rich actionable data dashboards, providing operators graphical notice of conditions in time for them to react. Engineers and managers are mining the database using graphical tools to visualize a vast amount of data easily, and then providing proactive process improvement measures in real time or generating automated or process-specific reports when desired – without IT intervention. And, of course, operators in the factory are given the information and guidance they need to do the best possible job.
Consolidation of disparate software and information systems in the factory is becoming a necessity. Consider how traceability, reporting and analysis demands are increasing, as is the need to reduce costs. These two demands appear to run counter to one another; historically, this has been true because of the inefficient way manufacturing information is managed. With the Lean philosophy driving the demand for comprehensive, operations-focused MOS software systems, the personnel and IT overhead required to achieve world-class process speed, control and visibility is greatly reduced, while at the same time management is provided with the data and measurements necessary to continually reduce or remove waste and inefficiencies within their enterprise.
Click here for the illustrations: Figure 1 |
