Data collection is no longer a back-office task. It is now the foundation of manufacturing intelligence. In this episode of Manufacturing Hub, we are joined by Brian Bribe, founder of Mach Controls, to explore the practical realities of modernizing data infrastructure inside manufacturing facilities.
Brian brings a frontline perspective to OT architecture and walks us through what it actually takes to connect legacy equipment, build scalable pipelines, and enable true real-time decision-making. We dive deep into Unified Namespace (UNS) principles, the evolution of MQTT and Kafka in industrial settings, and why so many manufacturers struggle to get ROI from new digital systems. From co-op student to founder of a systems integration firm, Brian shares how his early hands-on experience shaped his understanding of controls, business systems, and the gaps in between.
Topics include historian layers, challenges with SCADA-based centralization, how to scope a machine connectivity project, practical change management tips, and the path to flattened architectures using modern pub-sub tools. This is a must-watch for engineers, plant managers, and decision-makers looking to make sense of the data revolution inside their factories.
📌 Episode 217 is the first in our July theme focused on Data Collection for Manufacturing Intelligence.
🔗 Learn more about Manufacturing Hub: https://manufacturinghub.live
🔗 Explore Joltek consulting services: https://www.joltek.com/
⏱️ Timestamps
00:00 Introduction and Overview of Data Collection Theme
02:00 Brian’s Journey into Controls Engineering and Integration
06:30 What Legacy Manufacturing Equipment Actually Looks Like
10:00 Why Data Collection is Often an Afterthought in Scope
13:50 The Disconnect Between Machines and Business Systems
17:00 Real-World Discovery Process for Machine Connectivity
22:00 Overcoming Internal Resistance to Data-Driven Projects
28:00 Strategies for Extracting Data from Obsolete PLCs
34:30 Direct Sensor to SCADA vs PLC-Driven Design Debate
41:00 Flattening the Architecture and Breaking the Purdue Model
55:00 What Unified Namespace Looks Like in Practice with UMH