MESA Knows Smart Manufacturing

Following on the heels of a great community discussion, I have Joanne Friedmann and Khris Kammer discuss the elevation of data to the forefront of IT/OT architecture. Embedded in that is a thinking about the value of data and how it be measured, your Return on Data. Smart manufacturing and the MESA model place data at the foundation of any effective architecture. There is a significant shift in IT/OT where data is becoming the core driver of architecture decisions, rather than just an afterthought. Traditional IT/OT architectures were built around applications, with data being managed as a byproduct. Now, data-first architectures ensure that storage, processing, and movement of data are optimized for real-time decision-making, analytics, and AI-driven automation. AI and machine learning models thrive on high-quality, well-structured data. Data must be easily accessible across different systems (ERP, MES, SCADA, etc.). Organizations are implementing unified data governance frameworks to ensure consistency, security, and compliance. Modern architectures use event-driven models to react in real-time, which is critical for smart manufacturing and predictive maintenance. Manufacturing execution system (MES) platforms will evolve to be data-first, prioritizing real-time analytics. AI-driven decision-making will require IT systems to be built around data streams, not just applications. Data interoperability between legacy and modern systems will be crucial.

The large language model (think Chat GPT) made quite a splash at the end of 2023. This class of technologies can be applied to many real-world challenges that face industrial enterprises. But what problems are practical to be solved today by the application of LLM technology? The MESA Analytics Working Group discusses the challenges and opportunities that LLM bring to manufacturing. Join us, its a great discussion filled with possibilities for a promising new technology.

Why do 87% of data science projects fail? Non-Availability of Quality Data .... The data is usually raw and may contain many missing or absurd values. In such cases, it sometimes becomes impossible to make the given dataset into a model-friendly dataset. Thus, if the data quality is not good enough, the data science project will likely fail. What is the anecdote? ... Data Governance. But the truth is that data governance is a big challenge for each enterprise. According to a Gartner survey, over 90% of data governance projects fail to perform well. The MESA Analytics Working Group explores what good data governance could be and how it can reasonably be achieved.

How will industry jobs be affected by the increasing application and integration of analytics. With all the buzz around new technologies in AI/ML, ChatGPT, etc. this topic is “ever” relevant. Everyone is speculating and considering what this impact will be (some interesting links below) ⁠https://www.whitehouse.gov/wp-content/uploads/2022/12/TTC-EC-CEA-AI-Report-12052022-1.pdf⁠ ⁠https://www.mckinsey.com/featured-insights/future-of-work/ai-automation-and-the-future-of-work-ten-things-to-solve-for⁠ ⁠https://www.wsj.com/articles/how-ai-change-workplace-af2162ee?mod=Searchresults_pos4&page=1⁠ ⁠https://www2.deloitte.com/us/en/insights/focus/technology-and-the-future-of-work.html⁠ In fact, what does Chat GPT think.... The increasing application and integration of analytics in manufacturing will have a significant impact on manufacturing jobs. Here are some ways in which manufacturing jobs may be affected: Automation and Robotics: Analytics can be used to optimize production processes, identify bottlenecks, and improve efficiency. This often leads to the implementation of automation and robotics technologies in manufacturing facilities. As a result, certain manual and repetitive tasks previously performed by humans may be automated, reducing the need for labor in those areas. Upskilling and Reskilling: With the integration of analytics, manufacturing jobs will require a higher level of technical proficiency. Workers will need to acquire new skills to effectively operate and maintain advanced machinery, analyze data, and interpret insights derived from analytics platforms. Upskilling and reskilling programs will become essential to ensure the existing workforce remains relevant and adaptable. Data Analysts and Data Scientists: The integration of analytics in manufacturing will create a demand for professionals skilled in data analysis and data science. Manufacturers will need experts who can collect, analyze, and interpret large volumes of data generated by various systems, such as sensors, Internet of Things (IoT) devices, and production equipment. Data analysts and data scientists will play a crucial role in optimizing processes, predicting maintenance needs, and making data-driven decisions. Quality Control and Predictive Maintenance: Analytics can improve quality control processes by monitoring production data in real-time, detecting anomalies, and identifying potential defects early in the manufacturing process. This can lead to a reduction in the number of manual inspections required, but it will also create a need for skilled technicians who can oversee and maintain the analytics systems used for quality control and predictive maintenance. Decision Support Systems: Analytics can provide valuable insights to support decision-making in manufacturing, such as optimizing inventory levels, forecasting demand, and identifying cost-saving opportunities. This can lead to more efficient resource allocation and strategic decision-making. However, decision support systems may also result in a shift in job roles, with a greater emphasis on data-driven decision-making and a reduced need for manual planning and forecasting. It is important to note that while some manufacturing jobs may be affected or replaced by automation and analytics, new job opportunities will also emerge as companies adapt to these technologies. Workers with the ability to embrace and leverage analytics, as well as those involved in designing, implementing, and maintaining the analytics systems themselves, will likely find new avenues for employment in the evolving manufacturing landscape.

ChatGPT, the Chat Generative Pre-trained Transformer, is all the rage.  Speculation abounds on the technologies applications and who might be de-careered because of it.    The analytics working group at MESA has an informative discussion on what ChatGPT is today and how technology like ChatGPT may be applied across manufacturing, especially when challenged with synthesizing massive amounts of industrial data.  As a natural language user interface, tools like ChatGPT may help enable collaborative knowledge management systems and interactive, prescriptive analytics.

In data analytics better than half the time in the project effort may be spent on data acquisition and preparation.  Capturing meaningful and accurate manufacturing metrics for driving data analytics can be very difficult.  But when there is a high level of uncertainty about a measurement, we need to consider how much accuracy we truly need to obtain meaningful insights from our analytics.   The true goal of any measurement is not perfection but to reduce uncertainty. The podcast crew of the MESA Analytics Working Group is joined by guest Kip R. Krumwiede to discuss how to approach measurements that provide what we need to satisfy our analytic end goals.

The analytics working group at MESA International discusses how to discuss, scope, prioritize and measure the financial benefits when justifying Smart Manufacturing concepts to C-level organization executives.