Five smart manufacturing hurdles to overcome

Smart manufacturing is designed to rapidly create applications that enable collaboration among all people, systems, and assets, but there are key hurdles that need to be addressed for this to happen.

By Brad Williams June 23, 2016

Manufacturing generates more data than any other industry. New levels of connectivity, advanced computing, smarter sensors and devices, and improved data access and storage are increasing the breadth, volume, and resolution of available data. If you just listen to the Big Data hype, the assumption is that business value can be derived if data can be harnessed. In the future, manufacturing companies will gather more data, but will it be used effectively? That’s the key question that needs to be addressed, and smart manufacturing is working on answering that.

Smart manufacturing goals

The overall goal of smart manufacturing is to rapidly create applications that enable collaboration among all people, systems, and assets across a value chain-applications and architecture that build a new smart manufacturing software platform for the future.

Manufacturing enterprises know that access to data is important for improving operational performance. Performance improves when contextual information is provided at the right time to the right role to enable decision-making.

New technologies are making more data available every day. Some manufacturers are applying big data and analytics technologies in the hope that intelligence mined from this data will enable them to reach new levels of operational performance.

Applications are key to smart manufacturing

Success is not merely about data acquisition and providing visibility to more detailed and diverse data. In fact, it is not just about data. It is about rapid creation and delivery of rich interactive applications. Examples include:

  • Applications that provide just the right amount of the right information to the right role at the right time to allow them to make the right decision and take immediate action.
  • Applications that communicate between systems driving workflows and autonomous action.
  • Applications that prevent and predict issues by applying artificial intelligence and machine learning technology.

In other words, it is not just pumping out big data, it is about applications using and providing "right data."

Smart manufacturing applications are essential for extracting value from data and they are role based and real-time. They also address specific use cases aligned with unique business challenges. The applications provide proactive notifications to enable rapid resolution of issues. They are deployed to diverse form factors ranging from wearables to mobile devices to industrial touchscreen PCs.

The applications may merge the physical and digital worlds using augmented reality. The right applications are agile, and they quickly and easily change with the business. Smart manufacturing requires the right applications. But there are obstacles to delivering smart manufacturing applications.

Five smart manufacturing hurdles 

1. Data contextualization

Enterprises want to make intelligent decisions from data. But, an enterprise typically has many diverse software systems so techniques for obtaining data from these systems vary. It can also be difficult to combine business system data with manufacturing process data, yet data contextualization is essential to make intelligent decisions.

2. Underfunded and overwhelmed IT

Internal information technology (IT) organizations are often underfunded and overwhelmed. They do not have the capacity to deliver the applications needed and in the future there will be a demand for more applications and application management. Optimally, application composition should be democratized so that it can enable business users to make smart decisions.

3. Change and change management

Today, change is the norm. So applications must be agile because the business strategy, products, customer demands, and IT systems are constantly changing. These challenges will be compounded in the future because advancements in technology will drive a massive increase in the amount of available data resulting in demands for more applications.

4. Legacy software

Even if the applications are built, conventional software is not sufficient to meet the scale and diversity of applications required. Most manufacturing personnel are often given access to read-only dashboards providing a rearview mirror perspective on performance and offering no means of issue resolution or avoidance.

Also, there are issues within the two approaches typically applied to applications: IT and operations technology (OT).

Applications supporting manufacturing must span these approaches (IT and OT). Continuous improvement inherently necessitates agile applications. But, legacy systems are often difficult to extend and support. They cannot be quickly adapted to align with evolving needs.

5. Disparate systems

Existing interfaces between disparate systems can be problematic. Current approaches center on systems of record and on moving data between systems. As the volume of data increases this approach becomes more complex and less practical.

A better way

A new approach is needed to reach the goal of smart manufacturing. Luckily, solutions can be built without replacing legacy systems. In fact, we can use the same modern software technologies we use in our personal lives to build new composite, role-based applications that can span legacy systems and new sources of data. Modern software technologies include cloud, mobile, big data, social, and advanced analytics technologies which are robust and proven.

Brad Williams is a consultant at ThingWorx. He is also a MESA International smart manufacturing working group member. This article originally appeared on MESA International’s blog. MESA International is a CFE Media content partner. Edited by Chris Vavra, production editor, CFE Media, cvavra@cfemedia.com.

ONLINE extra

See additional stories about smart manufacturing linked below.

Original content can be found at blog.mesa.org.