Walk the Floor: Why Your Asset Count Is Wrong

Work with Aaron: https://protectitallpod.com/work/ The book: https://protectitallpod.com/book/ This episode: https://protectitallpod.com/ep131/ The OT Security Starter Kit: https://protectitallpod.com/starter-kit/ Kevin Kumpf started out strapping steel under gantry cranes, found the crane control systems had never been updated, and has worked in OT security ever since, including leading OT security at a large US utility. He and Aaron spend this episode on the things that do not show up in a tool: safety culture, knowing what good looks like, and walking the floor. Why an IT data center would never post a "days since last incident" sign, and why OT treats safety as the job. Two engineers at one plant who count 50 and 200 control cabinets. Ten engines on one ship, one running different. Why plants protect the most expensive asset instead of the most critical one, and the five-cent part nobody budgets for. Then data and AI: a boiler tube failure spreadsheet that decided outage plans without ever being on a critical asset list, why AI without context is just numbers, and why a human who can check the work still matters. Aaron's zip tie that left a core switch failed for months, the only backup server sitting next to a space heater, and how to sell OT security to the business: uptime, reliability and competitive advantage, done securely. Key takeaways Count it yourself. If two people at the same plant give you 50 and 200, you do not have an inventory, and a tool will not settle it. Know what good looks like. A baseline of temperatures, flows, RPMs and who talks to whom is what makes the odd one out visible. Criticality is not price. Ask what stops production, including the small part and the small site on the same network. Walk the floor and ask what the red light means. If nobody knows, find out or remove the light. AI needs context and a human who can check its answer. Lead with uptime and reliability. Security comes along with it. Chapters 0:00 Intro 0:22 Welcome and introduction 1:58 From strapping steel to crane controls 3:16 Hard hats and safety culture in OT 6:08 Days since last incident: IT vs OT 6:58 A sticker for every year without an incident 8:29 No shortcuts to safety 9:54 Teaching IT people the safety mindset 13:11 No-judgment zone: who knows their baseline? 15:39 50 cabinets or 200: asset counts that disagree 17:17 Who controls the plant, and knowing your baseline 21:17 You have to know what good looks like 22:41 Ten engines, one running different 23:49 The most expensive asset is not the most critical 25:03 The boiler tube failure spreadsheet 28:04 Excel, history and data in context 30:57 If the AI says 42 33:42 Data versus information 36:17 The cloud and trust but verify 40:47 Firewalls, passive monitoring, people and process 42:41 Enterprise IT tools in a plant, done the OT way 44:58 One zip tie and a failed core switch 48:03 The backup server next to the space heater 50:48 Protecting the device from the human 51:15 What Kevin is building ... Chapters (00:00:00) - Introduction and Guest Welcome(00:01:03) - Kevin's Origin Story in OT/Steel Plant(00:03:13) - Safety Culture in OT vs IT(00:08:30) - Safety Incident Stories and Lessons(00:10:24) - Building Safety Mindset in Teams and Conferences(00:13:06) - Not Knowing Your Asset Baseline(00:17:09) - The Refrigerator Hack and Monitoring Failures(00:21:17) - Knowing What Good Looks Like - Baselines(00:23:39) - Protecting the Wrong Assets - Critical vs Expensive(00:25:18) - Boiler Tube Failure Report - Hidden Critical Processes(00:27:51) - AI, Context, and Data vs Information(00:31:12) - Human in the Loop - Verifying AI and Data(00:33:39) - Data Monetization and Cloud Risks(00:37:16) - Trust But Verify in OT Practices(00:40:31) - Vendor Overpromises and Tool Limitations(00:42:42) - Enterprise IT Tools Applied to OT