Considering the Adoption of AI Agents for Machine Tooling: Pros and Cons Worth Pondering
Rewritten Article:
Take a peek into the future where your factory tools are equipped with their own representatives - Agentic AI. Just like a Hollywood agent, it handles the nitty-gritty tasks, allowing you to focus on the big picture.
Intrigued? Imagine your lathe having an AI that adjusts its spindle speed and workpiece feeding according to dynamic conditions. It's not just a simple on/off switch; it can slow down when temperatures grow hot or predict failure due to vibration analysis - all while working alongside the maintenance AI to schedule upkeep.
Though not imminent, this vision is becoming more tangible with time. If manufacturers incorporate this tech in a significant way, the roles and responsibilities in the industry could undergo a massive shift.
Last week, at the Xcelerate 2025 conference, I had the opportunity to chat with Aaron Merkin, CTO at Fluke Reliability, about Agentic AI and its potential impact on the plant floor. Let's dive in.
Aaron Merkin: Typically, process control systems follow a "define-the-process-flow" approach. The exciting part is that Agentic AI could enable individual nodes to act more autonomously and adopt a choreographed model. This allows for much more flexible and responsive operations.
Think about a factory scenario where a machine adjusts its speed when a fault condition arises. Traditionally, an operator would detect this condition and respond accordingly. In an Agentic AI setting, the machinery would itself recognize the slowdown and adjust its own speed without human intervention.
Dennis Scimeca: If each station on a production line can control itself and read the conditions before and after, wouldn't that eliminate the human factor entirely?
Let's say we have someone monitoring the entire process on the floor. If Agentic AI decides to slow down a machine, it's already made the decision. A human would still be needed to validate these decisions, but rolling that human into every minor correction would defeat the purpose of using AI for faster decision-making. So where does the human come into play?
AM: In traditional process control, humans serve as monitors and responders, making adjustments when necessary. Even with Agentic AI, humans will still have a role in setting the boundaries for the specific decisions the autonomous nodes can make.
That's where human expertise comes into play...tuning the Agentic AI behaviors based on human understanding of what can and cannot go wrong.
DS: Manufacturing leaders are constantly concerned about the shortage of skilled workers as veterans retire. If a departing operator tweaks the Agentic AI to perfection, couldn't that be considered knowledge transfer? And doesn't that bring up concerns about AI replacing people?
Some parameters will still change, but eventually, Agentic AI could reach a critical mass of knowledge where it can adapt to fresh changes without human tweaks. The AI could even self-correct based on its experiences.
AM: Unfortunately, the shift could lead to the loss of low- to medium-skilled positions. Basic data collection and trend analysis jobs will be automated. But higher-level analysis will still require senior experts evaluating the AI's decisions.
There is a balancing act between the growing need for lower-skilled operators and the increasing demand for highly skilled experts to oversee the AI.
Long-term, I am concerned that the industry will focus on replacing low-skilled workers. However, we will face an expertise gap in the future as we have not trained enough people to become the experts who oversee these systems. This gap may not become apparent for another generation.
DS: In a world where low-skilled jobs are replaced by either cobots or AI for basic analysis tasks, what will starting positions at manufacturing companies look like?
AM: In a practical sense, factories will prioritize soft skills for automation. Highly skilled jobs will remain a necessity for overseeing the AI's performance. To lure new generations into manufacturing, manufacturers need to reconsider their approach and promote skilled trades as an attractive, financially rewarding career path.
Phew! That was a short glimpse into the Agentic AI-powered future of manufacturing. Buckle up, because we're in for one heck of a ride!
Enrichment Data:
Overall, Agentic AI is set to revolutionize manufacturing, moving from static automation to adaptive, autonomous systems that solve problems proactively, optimize operations in real-time, and digitize workforce knowledge. The core impact areas include:
- Proactive problem-solving: Agentic AI detects issues (vibrations, temperature fluctuations) and schedules maintenance, orders parts, or pauses production to prevent downtime.
- Real-time optimization: AI analyzes data from sensors and external factors (weather, order urgency) to adjust production workflows dynamically.
- Digital twins: Virtual replicas of manufacturing processes allow for risk-free scenario testing, such as supply chain adjustments or energy consumption modeling.
Agentic AI doesn't wipe out human workers but fundamentally redefines their roles, focusing on exception management, strategic planning, and innovation in problem-solving. As AI continues to mature, humans will retain their crucial roles in creativity, ethical oversight, and complex decision-making where intuition remains crucial.
- In the future, Agentic AI could handle the intricate tasks of adjusting spindle speeds and workpiece feeding in manufacturing, akin to a Hollywood agent managing its client's careers.
- With Agentic AI, individual workstations on a production line may operate more autonomously, making decisions such as speed adjustments based on factors like vibration analysis and temperature fluctuations.
- While AI could take over some low- to medium-skilled jobs in manufacturing, there will still be a need for higher-level experts to oversee the performance of the AI, particularly in the areas of exception management, strategic planning, and innovation.
- The manufacturing industry, as it transitions to Agentic AI, could see a shift in roles and responsibilities, leading to a demand for reconsidering the approach to attract new generations, prioritizing soft skills, and promoting skilled trades as financially rewarding, future-proof career paths.
