AI, Automation & Human Judgment: Building Smarter Manufacturing Systems with Darragh de Standún
From The Manufacturers Network · 29:58
Show by Lisa Ryan
TL;DRAI and automation are changing manufacturing, but the smartest systems still need people.Darragh de Standún, CEO of Automated Industrial Robotics (AIR), joins Lisa Ryan to discuss how manufacturers can design automation around the people, products, and processes it needs to support. They explore AI-powered vision systems, human judgment, collaborative engineering, reshoring, automation strategy, and why manufacturers should involve automation partners before expensive design decisions get locked in.Darragh also shares how AIR brought six automation companies together into a single integrated organization while preserving the culture and expertise that made those businesses valuable in the first place.The big takeaway: The goal of manufacturing automation shouldn't simply be replacing people. It should give people better information, better tools, and better systems for decision-making.Episode DescriptionWhen manufacturers talk about AI and automation, the conversation often jumps immediately to replacing people.Darragh de Standún sees it differently.As CEO of Automated Industrial Robotics (AIR), Darragh leads an organization of roughly 600 people created by bringing together six automation companies across the United States, Ireland, and England.In this episode of The Manufacturers Network Podcast, Lisa Ryan talks with Darragh about what effective industrial automation actually looks like, from the earliest stages of manufacturing system design to AI-powered quality inspection and the decisions that still require human judgment.Design for Automation Before the Expensive Decisions Are MadeOne of the biggest automation mistakes can happen long before a robot ever reaches the plant floor.Product tolerances, component design, assembly sequence, tooling, production volume, and even small design choices can dramatically affect how easily a product can be automated.Darragh explains why manufacturers should involve automation expertise early in the design of products and manufacturing systems.A radius here or a tighter tolerance there may look like a small engineering decision. Once production begins, it can make an enormous difference in whether an automated system operates reliably.The earlier manufacturing and automation are designed together, the more opportunity there is to avoid expensive fixes later.What Collaborative Engineering Looks LikeAutomation can't be designed in isolation from the product it's supposed to manufacture.Darragh describes AIR's approach to collaborative engineering, where engineers examine the product, subcomponents, tolerances, assembly sequence, testing requirements, production goals, and available technologies together.Instead of simply building whatever appears in a specification, the team asks whether the sequence should change, whether tolerances make sense, and which automation technology is appropriate for the job.That might include six-axis robots, SCARA robots, pick-and-place systems, vision systems, or a combination of automated and operator-driven processes.The objective is to create a manufacturing system that aligns with the product's reality.How AI Vision Systems Are Changing Manufacturing QualityTraditional machine vision required manufacturers to anticipate defects.Engineers needed examples of known defective parts so the vision system could be trained to reject them. Sometimes that meant deliberately recreating scratches, surface problems, or other defects simply so the system could recognize them later.Machine learning changes that equation.Darragh explains how newer AI-enabled vision systems can learn what constitutes a good part and identify deviations that warrant attention.When the system encounters something outside its expected parameters, it can quarantine the part and flag it for human review.An engineer, technician, or operator can then determine whether the part is actually defective. That human decision becomes additional information the system can learn from.Over time, the system becomes better at recognizing defects without requiring engineers to anticipate every possible failure mode.Where Does Human Judgment Fit Into AI and Automation?This may be the most important question in the episode.Experienced manufacturing employees develop knowledge that doesn't always live in a manual, database, or dashboard.Sometimes an operator who's spent 20 years around a process simply knows something is off.So what happens when the dashboard says everything is fine, but the experienced person standing beside the machine disagrees?Darragh explains that AIR approaches automation by asking how technology can enable people to perform better.The goal is to give operators better information so they can make faster, easier, and more informed decisions.Machines and people aren't separate systems. The opportunity is to improve how humans interact with machines.As AI becomes more capable, preserving that human judgment becomes increasingly important.Building Automation…
More episodes of The Manufacturers Network
More episodes
- Manufacturing Leadership: How Culture, Accountability, and AI Drive Growth with Tim Beatty
- Why Manufacturers Can’t Find Workers: Fixing the Frontline Recruiting Process with Sam Beiler
- You Bought It. Why Aren’t They Using It? with Rich Thomas
- When the Data Says One Thing and Your Operator Says Another with Michael Dorman
- Industry 4.0: How Manufacturers Can Combine AI, Lean and Human Expertise with Mike Ungar
- Manufacturing Economic Outlook: Recession Risk, Tariffs, Labor, AI, and Supply Chains with Bill Conerly
- Manufacturing Traceability: Fighting Counterfeit Parts with AI | Roei Ganzarski
- Manufacturing Insurance: The Coverage Gaps That Can Cost You with Michael Senderovich
- Manufacturing Leadership: Why Being Right Isn’t Enough with Jared Scott
- Leadership Under Pressure: Burnout, Communication & Human Judgment in Manufacturing with Kabrina Ashley
PodSearch is a free, independent podcast search engine. Read about PodSearch, the privacy policy or contact us.