On May 29, TMEP hosted the CESMII IT/OT Convergence Smart Manufacturing Roadshow, its first event to help manufacturers build a strong foundation for AI with system interoperability and data architecture. AI is among the top options manufacturers want to learn about as they try to solve chronic problems on their plant floors. To implement AI, manufacturers must develop a long-term AI strategy that builds upon their core competencies to strengthen their competitive advantage. In other words, they must apply their unique experience and knowledge of their core processes to AI, ie. machining, cutting, bending, welding, assembling painting, etc. Manufacturers must build their own internal capability to collect data and use advanced analytics tools.
The event hosted the Clean Energy Smart Manufacturing Innovation Institute or CESMII. CESMII presented an overarching view of the tools that are needed to fully connect people, materials and processes across the enterprise. Chief Technology Officer Jonathan Wise articulated the need for graph databases and the use of unified data architectures. These are the building blocks of the future. In addition to AI applications, system interoperability will drastically reduce integration costs for manufacturers by lowering the hurdles that are needed to bring them online.
Currently, manufacturers must outsource AI to vendors who take a small data set from a client, perhaps from one or two machines. The vendor runs the data through custom software and returns answers to clients. There are two fundamental problems with these types of projects:
- There is no guarantee that the data available is the data that will reveal a problem. A few companies have success stories of projects that save millions of dollars, but many AI projects fail completely or have very little payback. Failed AI projects are frequent because most companies have silo’d data systems that limit data sets for AI to run on.
- The projects do not engage the workforce. AI consulting companies have proprietary platforms, so the system is a block box. There is no opportunity to experiment and learn in an iterative fashion.
The complexity and scope of data that is needed to tackle manufacturing problems cannot be understated. The data needed likely does not reside inside one machine or software system. Many factors may lie within material, post machining assembly, or even the supply chain. The only way to guarantee success in applying AI is to create a sustainable long-term plan to build the internal capacity within the existing and future workforce to collect and analyze the data. In short, manufacturers must undergo a digital transformation.
The CESMII IT/OT Convergence event created a sustainable foundation for solving real problems with data and a sound fundamental understanding of manufacturing processes. The goal is to create a future workforce who is ready to be a part of digital transformation. We believe that predictive quality (and other business goals) can only be achieved by creating this solid data foundation and workforce engagement.
TMEP also hosted presentations that showed manufacturers how to envision a new era of data management from Zygo Corporation, ThinkIQ, Amatrol, Full Bore Digital and Tooling U-SME.
The University of Tennessee seeks to support manufacturers use of I4.0 and AI projects grounded with three core principles:
- a solid data architecture to connect silo’d software systems and databases.
- a firm understanding of the physics manufacturing processes
- a commitment to sound scientific research and development
Adhering to these principles and building upon the strong partnerships with CESMII, Zygo, and Tooling U, the University of Tennessee Center for Industrial Services and it’s partners at TN-MADE, the University of Tennessee Tickle College of Engineering, and the national NIST Hollings Manufacturing Extension Partnership, will provide unique capabilities for manufacturers to begin their I4.0 journey.
If you would like to stay informed about future events like this, please contact Danny Norman at danny.norman@tennessee.edu.
