Claude Code, Ignition SCADA, and industrial AI are changing how automation projects get built. Jonathan Hottell of Proton Automation joins host Mike Flores at ENTELEC to discuss his experience building SCADA systems with Claude Code and developing an MCP server for Ignition.
They explore why MQTT, open standards, and a unified namespace matter for bringing context to industrial data—and why that foundation matters for AI readiness in oil and gas. Jonathan also shares what changed his mind about AI and why faster development still requires engineering judgment and production testing.
The conversation covers:
• Using Claude Code and MCP with Ignition
• MQTT adoption, legacy RTUs, and vendor lock-in
• Unified namespace and contextualized SCADA data
• AI’s role in supporting SCADA developers
• The gap between a fast prototype and a production-ready system
• Potential applications in alarm management and production forecasting
• Local AI models, data security, and edge computing
• A preview of Jonathan’s Ignition Oil and Gas Collective presentation
Subscribe to the Energy Tech Podcast for conversations with the people building and using technology across oil and gas.
Where would AI make the biggest difference in your SCADA workflow: development, alarm management, or data analysis? Let us know in the comments.
CHAPTERS
00:00 Welcome and Jonathan Hottell introduction
01:13 MQTT adoption and legacy RTUs
03:08 Open standards and vendor lock-in
04:28 Unified namespace and AI-ready data
06:53 AI agents and the pace of change
07:42 Ignition MCP and building SCADA with Claude Code
09:06 Will AI replace SCADA developers?
12:20 From AI prototype to production-ready software
12:58 Ignition Oil and Gas Collective preview
15:40 Industrial AI use cases and sandbox testing
17:52 Local AI models and data security
20:41 Data orchestration, FlowFuse, and Node-RED
22:04 Edge computing and smaller AI models
23:47 OpSite alarm management and closing thoughts
Music: “Fast Life” by Uygar Duzgun
Courtesy of Epidemic Sound
#SCADA #Ignition #IndustrialAI
