This episode is brought to you by Cirrus Link Solutions and Canary Labs.
Cirrus Link Solutions - MQTT and Sparkplug software that connects OT to IT and cloud systems.
Learn more: https://unpluggediiot.com/go/cirrus-link
---
Why do things break, and how do you make sure they don't break the same way twice?
Peter Horsburgh, Managing Director of Sologic Australasia and an industrial reliability engineer with 20+ years across mining, power generation and smelting, joins Phil Seboa and Ed Fuentes to talk about root cause analysis. He explains why we jump to conclusions, why problem solving works like balancing an equation, and how low-cost industrial IoT provides the evidence an RCA needs. From a fifty-dollar sensor on a hydraulic pack nobody had seen running since 1983, to eight hours' warning of a failure on a V16 mining truck engine, Peter shows how evidence turns a guess into a permanent fix.
Topics covered:
- Why engineers jump to conclusions, and why diverse teams find more causes
- The five whys: when it works and when to step up
- Problem solving as balancing an equation: the spilled-water example
- Using cheap IIoT sensors to bypass the control system and fill data gaps
- Turning an unreliable production process around with trigger points
- The ROI of RCA, the fifty-cent washer and the "cause plus two" rule
- Five steps to build an RCA program, and a smooth/clean/cool/dry sensor strategy
Guest:
- Peter Horsburgh - Managing Director, Sologic Australasia
- LinkedIn: https://www.linkedin.com/in/extraordinaryreliabilityengineer/
Resources:
- Sologic's free RCA ebooks: https://www.sologic.com/en-au/resources/ebooks
- Example RCA, West Point Treatment Plant sewage release: https://www.sologic.com/en-au/resources/example-problems/west-point-treatment-plant-sewage-released
Episode page: https://unpluggediiot.com/episodes/ep-63-root-cause-analysis-iiot-peter-horsburgh
Chapters:
00:00 Introduction and sponsors
00:36 Welcome and introducing Peter
01:58 Peter's path: mining, power gen and smelting
04:51 What Sologic does: navigate uncertainty with confidence
06:08 Why we jump to conclusions, and diverse teams
08:10 When RCA works: getting the CEO on board
09:27 The five whys and its limits
11:04 Who owns and leads an RCA
13:32 Why problem solving isn't well taught
16:33 How long does an RCA take?
17:33 Where IIoT fits: trigger points and missing data
20:23 Bypassing the control system with a $50 sensor
22:17 Eight hours' warning on a V16 truck engine
23:38 Trust and an evidence-based process
24:45 Balancing the equation: the spilled-water example
28:49 Using cheap IIoT to verify causes
29:51 Turning around an unreliable production process
35:27 Too much stock: redesigning the maintenance strategy
38:19 Set your triggers, stick to the process
39:48 Building the business case: ROI
40:44 The fifty-cent washer, human error and cause plus two
43:55 Wearables, geofencing and the hierarchy of controls
46:15 Why mixed teams find better solutions
48:32 Five steps to better RCA
52:21 The sprint technique
54:18 Going proactive: FMEA and RCM
55:48 Smooth, clean, cool, dry: an IIoT sensor strategy
58:15 What Peter is learning: aerobatics and LLMs for risk
1:02:37 Wrap-up and how to connect with Peter
---
Canary Labs - The industrial historian platform that captures your plant and process data the moment it is generated and makes it instantly accessible across your entire operation. Visit: https://unpluggediiot.com/go/canary-labs