John Clemons is a Solution Consultant for Rockwell Automation. He’s been working in the field of Manufacturing Technology for over 30 years.
Artificial intelligence (AI) is changing the way we live and work. From healthcare diagnostics and fraud detection to customer service chatbots, autonomous vehicles, facial recognition and virtual assistants like Siri and Alexa—AI is everywhere in our daily lives.
In manufacturing, AI has been a key technology for many years. But the type of AI traditionally used in manufacturing differs from the generative AI (GenAI) we’re becoming familiar with today. Traditional AI in manufacturing isn’t focused on generating new content. Instead, it takes large amounts of data, detects patterns in that data, applies a series of rules to the patterns and then makes predictions or recommendations based on those insights.
Meanwhile, GenAI takes large amounts of data as input and then generates specific outputs based on that data. For example, when we ask a virtual assistant a question, GenAI analyzes large amounts of data and generates a response.
Let’s take a closer look at how traditional AI and GenAI are used in manufacturing to gain a deeper understanding of their differences and explore examples of how they can work together to drive innovation and efficiency.
Traditional AI Vs. GenAI
• Purpose: Traditional AI in manufacturing detects patterns in data and makes predictions. Its goal is to understand what’s happening now and forecast what will happen next, so that the right decisions can be made. On the other hand, GenAI is designed to generate new content—training materials, documentation or new product ideas such as recipes.
• Functionality: Traditional AI’s underlying model is trained on large amounts of historical data, often using predefined rules and algorithms to detect patterns and make predictions. GenAI also uses large amounts of data, but instead of detecting patterns or making predictions, it uses that data to generate new content.
• Learning: Traditional AI learns from data and is often even trained to look for specific data patterns. It uses rules and algorithms to learn. Traditional AI optimizes for accuracy in its prediction and classification tasks. GenAI also learns almost exclusively from the data it’s trained on and adapts its behavior based on the content and input it receives. GenAI optimizes for generating realistic, contextually appropriate outputs.
• Adaptability: Traditional AI is trained for each new application, as it needs to detect patterns and learn the specific patterns it’s detecting. GenAI learns and adapts from the data it’s provided and even from the results it generated in the past. The mechanism that GenAI uses to be more adaptable is reinforcement learning—a traditional AI approach that’s being applied to GenAI. GenAI’s strength is in its flexible content generation.
Examples Of Traditional AI In Manufacturing
• Predictive Maintenance: Traditional AI analyzes large amounts of equipment data to detect operational patterns. It then predicts potential failures the equipment might have so that the proper maintenance can be performed at the right time.
• Vision-Based Quality Control: Traditional computer vision with rule-based classification uses traditional AI models to process data from vision-based quality inspections to determine pass/fail inspection outcomes, identify root causes of the defects and find the best course of action for correction. The latest vision systems increasingly use GenAI models in a hybrid approach with these traditional AI models.
• Production Scheduling: AI uses data on orders, constraints and operating conditions to create and update production schedules in real time as variables change.
• Sustainability: AI optimizes energy and utility usage while minimizing waste and emissions from the facility by analyzing data throughout the manufacturing facility, all in real time.
Examples Of GenAI In Manufacturing
• Training Materials: GenAI processes manuals, standard operating procedures (SOPs), engineering drawings and work instructions to generate new content such as training guides, exercises and certification materials.
• Shift Reports: GenAI collects data from production, quality and materials activities during a shift, then analyzes and summarizes it to generate the required shift reports and handover materials, highlighting the most significant items for the upcoming shift.
• Recipes And Specs: When new product variants are produced, GenAI takes existing recipes and specs, applies the required product changes and updates the documentation accordingly. It then disseminates those recipes and specs to the people and systems that need them.
• Quality Assurance (QA) Testing: When product tests fail, GenAI can create new test protocols to enable more detailed testing and help classify the problem or determine if a problem truly exists. If a problem is identified, GenAI can then generate the test protocols to ultimately determine the root cause and the correction. GenAI can also analyze failure patterns and generate hypotheses with customized testing procedures and then create detailed root cause analysis reports synthesizing data from multiple sources.
Working Together
At this point, you might think that traditional AI and GenAI both make a lot of sense and have distinct roles in manufacturing but that they don’t appear to intersect or really overlap much at all.
Nothing could be further from the truth. And yes, while they do each have their place and are very powerful in their own right, they both achieve their greatest value when they are working together.
For example, in equipment maintenance, traditional AI detects equipment anomalies, then GenAI generates contextualized maintenance instructions tailored to the technician’s skill level. In production scheduling, GenAI creates production schedule variations, and traditional AI evaluates which options best optimize efficiency.
By working together, they can produce tremendous value for manufacturing—much more than they can separately.
Conclusion
Throughout this discussion, we’ve explored how traditional AI and generative AI share similarities but differ fundamentally in how they operate and are optimized for different tasks, resulting in distinct objectives and outputs.
Part 2 will take a deeper dive into how traditional AI and GenAI work together to provide operators and maintenance personnel with exactly what they need when they need it. It will also highlight the added value they bring to digital twin technology, data analytics, sustainability, productivity and quality control.
By understanding how each type of AI functions—and how they complement one another—manufacturers can better leverage both to drive innovation, improve efficiency and adapt to changing demands.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
