Key Topics from This Episode
Eric Seme joined Tony on the Industrial Automation Roundtable for a wide-ranging conversation about how Fireball Industries approaches automation projects, what EmberNet is and how it works, and where AI and robotics are actually heading in manufacturing. The episode is aimed at business owners and operations leaders who are thinking about automation but haven't found a practical entry point.
The core philosophy Eric shares is to skip the multi-year master plan and find one high-value process to put online first. Most customers already know where their biggest problem is: a station with no visibility, a process that isn't hitting targets, a line that creates constant manual work. Start there, get data, get return on investment, then iterate. The technology moves fast enough that a two-year plan is usually outdated before it's finished.
EmberNet is explained as a real-time Linux operating system combined with a central web-based management plane and the ability to run any application payload: Windows VMs for legacy software like FactoryTalk View SE, modern SCADA like Ignition, or custom applications. The goal is to let the average controls engineer leverage world-class cloud infrastructure and technology with a few clicks, without requiring cloud connectivity to run.
On virtual PLC: EmberNet enables industry 5.0 control where a software PLC runs in a container on the edge node. Eric cites less than 100 microseconds of jitter with multiple virtual PLCs deployed per node, outperforming standalone hardware PLCs in some cases. A real customer example covered is a Fortune 500 manufacturer running automated equipment testing cells, transitioning from a legacy C codebase deployed manually to a containerized stack running Ignition and Coticus on a single real-time node.
The robotics and AI segments are grounded and practical. On humanoid robots: they're coming, but they're not production-ready yet, and safety and compliance requirements will take time to sort out. The bigger near-term gain is applying AI to existing industrial arms to make them more adaptive. On AI in manufacturing broadly: on-premises LLMs running on plant data to surface trends and allow natural-language interaction with production systems are where the real efficiency gains are coming from.
Topics Covered
How to start an automation project
Find the one area with the most pain and least visibility. Put that online first. Get data. Show ROI. Then expand. Don't build a master plan that takes two years to finish and is already outdated before you deploy it.
What EmberNet actually is
A real-time OS plus central web-based management plane plus application runtime. Deploy Windows VMs, containers, or custom software. Centrally manage everything across any number of nodes.
Virtual PLC and industry 5.0
Software PLCs running in containers on EmberRTOS achieve sub-100 microsecond jitter with multiple instances per node. Can replace traditional PLC hardware while adding fleet management, remote visibility, and containerized lifecycle control.