When we think of manufacturing, what comes to mind? A dark, dirty picture of dangerous factories filled with workers and machines repeatedly doing the same thing over and over again? Well think again!

The information technology (IT) revolution has finally come to factory floors around the world. Today, manufacturing is becoming highly-automated and IT-driven or simply put, "smart." Every day, advances in modern manufacturing technologies make factories smarter, safer and also more environmentally sustainable. Progressive businesses had strategically invested to transform their operations from cost centers into Smart Manufacturing Profit Centers that will dramatically increase sales.

Today, the U.S. has the world's oldest industrial base with plans to help businesses construct or modernize factories with 21st century smart manufacturing technologies. To encourage businesses to make such major long-term capital investments for "Smarter Factories", there is immediate need for an overall business climate that is within recent global manufacturing advancements. Moreover, there's a huge counter-intuitive benefit — as factories get smarter, they create an increasingly larger number of "indirect" jobs required to support them. Studies show that as factories get smarter, the employment multipliers will increase two, three, four or more times greater than the number of jobs directly in manufacturing. Modernization or construction of this next generation  "Smart Factories" can accelerate this growing trend and create entire new communities of very good, sustainable jobs that are "indirectly" necessary to support manufacturing.

Together with the technology and people benefits associated with "Smart Manufacturing", expanding the capability of the plant operating system is equally important as a key enabler in the construction of these next generation Smart Factories.

In any plant operating system, measurement is one of the key components for success. Maybe the most substantial or primary benefit of the "Smart Manufacturing" program is its ability to improve the measurement system. By having real time trends available online, operators and process control systems will have improved capability to make near real time decisions to better manage the process. This process optimization will reduce the number excursions, improve operational efficiency and provide better quality.

A second benefit that can be realized in an operating system that is enabled with "Smart Manufacturing" is the automation of KPI calculation, reporting and root cause analysis. In nearly every enterprise business there is strong interest in developing common KPI metrics for performance analysis and power of comparison.

Finally, a third operating system benefit that can be realized with the deployment of "Smart Manufacturing" is the ability to adopt and rapidly transfer "best practice" innovations across the enterprise. Within the "Smart Manufacturing" environment, it is very important to segregate ad-hoc development and experimentation from established enterprise standards. While it is critical for employees from all levels of the organization to innovate, the enterprise must recognize and approve "best practices" standards.  Once adopted, these standards will be propagated across the entire enterprise to quickly leverage the associated benefits to all operating locations.

"Investing in technology, equipment and automation" is one of the most important ways to increase competitiveness. While many manufacturers have historically asked U.S. federal policy makers in to "level the playing field" by lowering taxes, regulatory, health care and other costs, most companies have stayed competitive by continuously improving productivity to overcome the higher manufacturing costs.

For two decades, manufacturers repeatedly executed lean, Six Sigma, Kaizan and other cost reduction strategies that successfully streamlined operations and reduced waste.  U.S. Manufacturing has been one of the world's best productivity success stories.  However, those strategies are now reaching the law of diminishing returns. Investments in technology, equipment and automation are the next logical step to improve productivity and stay competitive. As the IT revolution hits modern factory floors, the next generation of 21st Century Smart Manufacturing will cause the equivalent of the Quality Movement and much more. Not only will it yield a new era of greater industrial productivity and global competitiveness, but these flexible factories of the future will be safer, cleaner, and more energy efficient (November 7th, 2012 - Posted by Smart Manufacturing).

Nearly every modern manufacturing facility has some level of process control and shop floor functionality implemented within the location. The gaps in current capability are not the result of a lack of measurement or control, but instead related to a limited understanding of process history, interdependencies and common definition. Without the ability to correlate data and model performance, efforts to continuously improve the overall process are not reaching their full potential.

It is the integration of the existing instrumentation, process control and shop floor systems that is the key technology enabler for the successful implementation of "Smart Manufacturing". This integration goes well beyond just simple communication between devices, and instead hinges on a comprehensive enterprise and facility data model. This data model stores historical performance, documents relationships and correlations, and facilitates common information rollup and aggregation. It is also a model that can easily be adapted to accommodate both location and enterprise information needs.

Technology can also be introduced to improve information consumption through the use of HMI devices and visualization tools. These types of solutions can help ensure that the right information is provided to the right people at the right time.

There are four primary technical components that need to be included in the "Smart Manufacturing" design. They are as follows:

• Process Data Collection and Storage (Historian)
• Common Manufacturing Execution Systems (MES)
• Manufacturing and Business Intelligence (MI / BI)
• Integrated Data Model

Each of these technical components need independent consideration based on the specific nature and current condition of the manufacturing enterprise or location.  These considerations are outlined below.

Process Data Collection and Storage (Historian)
The purpose of a historian is to collect detailed process data and store it into a time based historical database. Although there are a number of vendors who provide historian functionality, the two key technical aspects to consider when selecting a solution are its network connection capabilities and ease of data consumption.

The primary level of connectivity is to the process. This connectivity requires direct linkage to process equipment such as instrumentation, PLCs, HMIs, SCADAs, etc. The challenge in establishing this connectivity is the significant diversity that often exists in process equipment. In many enterprises, there are potentially hundreds of different devices that monitor and control the process. It is critical that the selected historian is able to connect to all of these different devices to enable a comprehensive "Smart Manufacturing" environment.

A second level of connectivity that is needed for the historian is the linkage to the MES and ERP transactional databases. This will allow for two way data interfacing between the process and shop floor, and also enable the automatic calculation and population ok KPI and ERP data.

In addition to connectivity, the historian also needs to allow for ease in the consumption of data. This not only includes the proprietary query and visualization tools that come with the historian product, but also its ability to allow generic integration through standard tools and programming environments. Because "Smart Manufacturing" needs to be capable of aggregating and analyzing data from many different sources, data access cannot be restricted or limited to only the proprietary toolset.

Common Manufacturing Execution Systems (MES)
Manufacturing Execution Systems (MES) have been recognized for many years as an essential toolset for managing operations on the shop floor. Although some manufacturers may suggest that they do not have an MES solution, nearly all use some form of people, paper, spreadsheets or applications to manage production from point of order entry to point of finished goods. These solutions include functions for scheduling, quality management, production reporting, equipment monitoring, etc.

The architecture for MES applications is generally standard in nature. It consists of a client-based application with a relatively standard transactional database backend. Nearly all MES applications need either significant configuration or development to meet the very specific shop floor needs of the manufacturer. The most critical design consideration that needs to be addressed is whether to buy a vendor supplied package solution or develop a custom application. Both approaches have their advantages, but the decision should really be based on the availability of a comprehensive solution that truly meets your needs. If your processes are fairly unique, sometimes the configuration of a packaged solution can be more difficult, costly, and less effective than a custom solution. In many cases, a hybrid model is the best solution. This approach utilizes vendor packages for common functions and allows custom development for unique functionality.

Whatever MES solution strategy is adopted, it is essential that there is strong integration with both the historian and ERP databases. The "Smart Manufacturing" concept relies on the automation of data summary, calculations and transfers. This minimizes the reliance on people and improves the accuracy and timeliness of reporting.

Manufacturing and Business Intelligence (MI / BI)
The most critical design considerations for establishing a manufacturing and business intelligence strategy are the ability to access multiple data sources and to present this information in a standard, non-proprietary toolset. The challenge is deciding exactly how to accomplish this objective.  With so many vendors advertising their own proprietary solutions, and a lack of understanding of manufacturing and business data structure, there are often many conflicting and confusing approaches that are being propagated throughout the enterprise.

Two steps need to be taken to overcome this challenge. First, a generic environment needs to be established for the presentation of dashboards and reports. Second, there needs to be an increased focus on the development of a comprehensive data model. It is not difficult to design a generic environment for the presentation of dashboards and reports. There are, however, certain criteria that are essential for success. First, the functionality should almost certainly be web based and allow access from many different devices including PCs, laptops, thin client, smart phones, etc. Second, it should also be easy to use (leveraging the data model) and allow end users to develop their own ad-hoc queries and analysis. Third, and perhaps most important, the environment should allow for the secure segregation of the standard "best practice" dashboards and reports from the individual ad-hoc analytics. Intelligence is a critical component of "Smart Manufacturing" and the concept cannot be a complete success without recognizing its value and importance to the enterprise.

Integrated Data Model
The final and perhaps most important, technical component that is required for a comprehensive "Smart Manufacturing" environment is an integrated data model. To be clear, the purpose of this model is not to replicate any data. Instead, it has three primary objectives. First, it is designed to provide a visual and logical structure with dynamic reference to the source data. Second, it is modeled to recognize the correlative relationships between many of the established data elements. And third, it is developed to provide automated information aggregation and rollup capabilities for the critical business KPIs (Key Performance Indicators).

The toolsets to design a visual and logical data model are relatively new in the marketplace. The most fundamental deliverable of the toolset is its ability to define data elements that are capable of dynamically referencing source information from multiple databases. In most cases, it is recommended to structure these data elements in an asset format. Each data element can then be defined as enterprise common, technology common or site specific. This helps ensure the solution meets the needs of the enterprise as well as those of the location. The visual and logical data model allows users to easily locate the data that is needed and to better comprehend the scope of the information that is being consumed.

An added benefit when establishing the visual and logical data structure is the ability to define some level of correlation among the many different data elements. As a starting point, perhaps the easiest correlation to define is that between the KPIs and the defined assets. By adding KPI elements to the logical data structure, these initial correlations can be established for further drill down analysis capabilities. When adding KPI correlation to the data model, it is also important to recognize the ability and benefits of automated data aggregation and rollup. All KPIs are based on some level of source data. Although there are certainly examples of manual entered source data, many of these elements are from an online source. In either scenario, however, it is critical to understand how that source data is calculated or rolled-up into a summary KPI. In some cases, these KPIs are even calculated and summarized in many different frequencies (i.e., daily, weekly, monthly). As a result, it is important to establish clear calculation and rollup procedures as part of the data model. By using an automated calculation process, the solution will ensure consistency and eliminate errors. It is also very important to recognize that the data model, as all other components of "Smart Manufacturing", is a living part of the environment. It will never be perfect, and will always be evolving. New data points and KPIs will be required over time, and changes will have to be effectively controlled to ensure the component maintains its integrity. A good change management process is certainly an essential requirement for the success of a "Smart Manufacturing" program.