Industrial Edge Controls

The machine changed. The standard didn't.

Critical systems inside regulated facilities have long operated on the same basic principles: a human is in control, there is proof, and important actions are recorded.

That was true for the assembly line. It was true for the PLC. It is true for AI.

The next evolution of the industrial controls stack is Physical Edge Infrastructure. It extends the proven hardware, safety systems, and control architecture that have governed industrial automation for decades.

Industrial Edge Controls Glossary

A — C — F — H — I — L — M — P — S — U

Air Gap

A system can be connected to other systems by wires, networks, or other communication paths. An air gap is a deliberate physical separation that prevents those systems from communicating or exchanging data across that boundary. Nothing crosses the gap unless a person physically connects the two sides. In an industrial facility, an air gap can separate a critical control system from outside networks, creating a boundary that cannot be crossed through software or a network connection.

AI Agent

An AI agent is a computer system that can see what is happening, figure out what needs to happen next, and take action. It can use information, tools, and the results of its own actions to decide what to do next. Unlike a system that follows a fixed list of steps, an AI agent can work through a problem and change its actions as the situation changes. In a factory, an AI agent might watch a machine, notice something changing, investigate it, and take action to keep the process moving toward its goal.

Chain of Custody

Every important action on a system needs to be traceable. The technician who opened the panel. The AI agent that changed a setting. The person who approved the change. Chain of custody is the record of what happened, when it happened, who or what did it, and who was responsible for the action. Each step connects to the next, so the history of a system can be followed from beginning to end. When AI acts, the chain includes the agent, the action it took, and the authority under which it acted.

Control Path

Every machine takes instructions from somewhere. A person gives a command. An AI agent makes a decision. A controller sends a signal. An actuator moves. The route from a decision to a physical action is the control path. It can pass through software, networks, controllers, switches, and actuators, with each part determining what can happen next. The control path is what connects a decision to the machine that carries it out.

Failsafe

Critical systems are designed to respond safely when something goes wrong. A loss of power, a broken signal, a failed component, or an unexpected condition can trigger a defined response. A valve may close. A brake may engage. A machine may stop. Another system may stay powered because losing power would create a greater hazard. The safe response is decided during the design of the system, based on what could happen when it fails. Failsafe means the system is designed so that a failure leads toward a known safe condition.

Failsafe Open / Failsafe Closed

When a valve or other final control element loses power, signal, or control, it can be designed to move to a known position. Failsafe closed means it moves closed. Failsafe open means it moves open. Which position is safe depends on the hazard the system is designed to control. A valve may close to stop a chemical feed, or open to keep cooling water flowing. The safe position is chosen during the design of the system based on what happens if control is lost.

Failsafe De-Energize / Failsafe Energize

When a control signal or power is lost, a system can be designed to become safe by removing energy or by maintaining energy. Failsafe de-energize means the safe action happens when power is removed, such as a brake engaging when current stops. Failsafe energize means the safe action requires power to be applied or maintained, such as a system that must keep a cooling function running to prevent a dangerous condition. The choice depends on the process, the hazard, and what the system needs to do when something fails.

Failsafe Human

When power, communication, control, or trusted autonomy is lost, a person can become the final means of putting the machine into a safe state. The person does not rely on the signal, the controller, or the screen. They can act directly on the physical system: the breaker, the valve, the switch, or the disconnect. Human is the safe state when the system is designed so that a person can take direct control and make the machine safe.

Failsafe AI

Some moments don't wait for a person to arrive. A pressure spike can happen in milliseconds. A dangerous condition can develop before anyone can reach the panel. Failsafe AI allows an AI agent to recognize the condition and take action when a human cannot respond fast enough. It acts within defined safety boundaries to move the system toward a safe state, while recording what it saw, what it decided, and what it did. When human control is restored, the person can review the action, take control, and decide what happens next. Failsafe AI does not replace Failsafe Human. It is the bridge between a dangerous moment and the person who ultimately takes control.

HMI (Human Machine Interface)

The HMI is where a person sees what a machine is doing and tells the system what to do. It can show temperatures, pressures, flow rates, alarms, and other conditions as they change. The operator uses the HMI to understand the process and give commands to the control system. The HMI is the interface, not the machine itself. Behind it are the controllers, networks, sensors, and physical equipment that make the action happen. The HMI gives the human a window into the process and a way to interact with it.

Human in the Loop

Every automated system has a cycle. A sensor reads a value. An AI agent or controller makes a decision. An actuator takes an action. The cycle repeats. Human in the Loop places a person at a defined point in that cycle, where they can review what the system is doing, approve or reject an action, change the decision, or take control. The system is designed to give the person a chance to act before, during, or after the automated action.

Human in Control

A human in control has the physical authority to decide what a system does. They can direct it, interrupt it, or stop it without depending on the AI or software to give them permission. That authority exists in the physical control path: a switch, breaker, disconnect, valve, or other control that the system cannot override from software alone. The person's identity and actions can be verified and recorded. AI can make decisions and take actions, but it cannot be the final authority over itself. Human control means a person has the physical means to take control away from the system.

Immutable Record

Every important action in a regulated facility needs a record. That record captures what happened, when it happened, and who or what took the action. An immutable record is protected so that what was recorded cannot be quietly changed, deleted, or rewritten later. A signature stays attached to the action. A timestamp stays with the event. The record can be checked against an independent source to show that it has not been altered. That is what makes the record trustworthy long after the action happened.

Lockout/Tagout (LOTO)

Machines can store energy even after they are turned off. Electrical charge, pressure, heat, moving parts, springs, gravity, and other energy sources can still cause harm. Lockout/Tagout is the process of isolating those energy sources before work begins. A worker identifies the sources, shuts them off, physically locks or tags the isolation points, and verifies that the machine cannot operate or release stored energy. The machine stays isolated until the authorized worker completes the work and the required steps are followed to return it to service.

Model Hardware Standard (MHS)

Every machine is different. A robotic arm. An oven. A mixer. MHS gives an AI agent a common way to discover a machine, understand what it can do, and operate it. It describes the machine's capabilities, physical characteristics, and safety limits, so an agent can work with equipment it has never seen before. Instead of building a new connection for every machine, MHS gives agents and machines a standard way to find each other, communicate, and work together.

Physical-Precision Data Hub

Every machine is constantly producing signals: vibration, pressure, temperature, sound, current, and much more. A Physical-Precision Data Hub gives AI agents a much more detailed view of what is happening in the physical world, including changes that people cannot see, hear, or feel. It captures the details and timing of those changes, so the system can understand what happened, what changed, and how different events are connected. When an AI agent can see more of the physical world than a person can, it can notice problems earlier, understand machines more deeply, and make decisions based on information that humans could never observe on their own.

PLC (Programmable Logic Controller) 

Every machine on a facility floor takes instructions from somewhere. Often, that something is a PLC. A PLC is a small industrial computer built to control machines. It reads signals from sensors such as temperature, pressure, position, and flow, follows programmed logic, and sends instructions to motors, valves, and other devices. It is built to run reliably in harsh environments: dust, vibration, heat, and electrical interference. It can run continuously for years. When an AI agent acts on a machine, the PLC or another industrial controller may be part of the path that turns that decision into physical action.

Proof of Control

Knowing who is in control is not enough. In a regulated facility, you need to be able to prove it. Proof of Control is evidence that a specific person had authority over a specific system at a specific time, and that their actions can be traced back to them. The record shows who acted, what they authorized or changed, and when it happened. That proof can include physical actions, digital records, and independent evidence from the system itself. If you cannot prove who had control, what they did, and when they did it, you cannot fully prove that control existed.

SCADA (Supervisory Control and Data Acquisition)

A facility can have dozens or hundreds of PLCs and other control systems, each running its own part of the process. SCADA brings information from across the facility into one place. It shows operators what is happening, raises alarms, stores information, and can allow authorized commands to be sent back to the control systems. A pressure spike. A temperature deviation. A valve that failed to open. In a large facility, the equipment being monitored can span miles. SCADA gives people a way to see and supervise the operation as a whole.

SIS (Safety Instrumented System)

Every facility has control systems that run normal operations. The SIS is not one of them. It is an independent safety system designed to detect specific dangerous conditions and take defined action to protect people, equipment, and the process. A reactor pressure climbing too high. A gas concentration reaching a dangerous threshold. A temperature rising beyond safe limits. When those conditions are detected, the SIS acts without waiting for the normal control system. Valves close. Feeds stop. Equipment shuts down. The SIS is designed to move the process to a safe state when normal controls are not enough.

Ultimate Authority

Every automated system operates inside defined limits. Verified inputs. Known conditions. Approved actions. Inside those limits, the system can act on its own. Outside them, it cannot decide what happens next. A condition appears that was not expected. A sensor no longer agrees with the process. A decision reaches beyond what the system was designed or authorized to handle. The system stops, holds, or escalates the decision. Ultimate Authority is the person or authority that has the final right to decide what happens next. The system must be designed so that the agent cannot make itself the final authority.

We also host events in the real world! Come say hi.