By Jag Gattu, CEO, UptimeAI
Analyst firms get first-hand insight into new technology innovations and have the advantage of seeing the full breadth of market offerings, without the bias of any one company or technology category. There are moments when these trusted third parties put language to something practitioners have been feeling but struggling to articulate.
When LNS Research published their latest Industrial AI Maret Landscape blog post, there was a new section of their market map—Agentic Operations. As I was reading through this latest take on the market, I recognized one of those moments.
How the Market Finally Caught Up
LNS Research has been researching and carving up different cross-sections of the industrial AI market since they first drew the line in the sand between Industrial AI and Advanced Industrial Analytics a few years back. In this early blog, they defined Industrial AI as:
Industrial AI is the application of Artificial Intelligence techniques, such as machine learning, deep learning, natural language processing, search, computer vision, causality, and robotics, in combination with industrial subject matter expertise, such as manufacturing operations fundamentals, heuristics, industry standards, asset failure libraries, process know-how, workforce-related information, etc., to solve critical problems for the manufacturing industry.
Some mid-2025 research sliced the market up by use case: Asset Monitoring, Process Analytics, Process Control, Productivity, and Safety. Without another natural home in this framework, agents were lumped in with Co-pilots in the productivity use case. But this came with the caveat that bigger things were on the horizon for agentic AI:
More recently, the focus has shifted toward agentic AI— leveraging AI not just as a tool for automating repetitive tasks but as a self-learning system capable of understanding and streamlining complex workflows and driving autonomous decision-making. While a few vendors have started to offer these agentic AI capabilities, there are still ways for the industry to have access to truly autonomous agents with these complex decision-making capabilities.
A common theme in the recent pieces from LNS Research—they don’t just describe where the market is, but also where it has to go. This isn’t an easy task when your market moves at, say, the speed of AI.
Defining the Next Category: Agentic Operations
In their March 2026 Industrial AI Market Landscape, LNS Research defined Agentic Operations as:
AI-native solutions focused on multi-agent orchestration, causal and reasoning systems to provide analytics intelligence and execution and control capabilities. These vendors are building decision intelligence from the ground up, using knowledge graphs, causal reasoning, and autonomous agents to move beyond dashboards and deliver measurable outcomes.
These are systems capable of analysis and execution — operating with intelligence and decision autonomy to drive actual operational outcomes. Not more dashboards, alerts, or insights. Systems that make decisions.
For years, the dominant story in industrial analytics has been about giving operators better information: detect anomalies earlier, predict failures further out, surface recommendations more clearly. All valuable. But it assumes the bottleneck in industrial operations is knowing what will happen.
The Decision Latency Bottleneck
Last fall, we were introduced to the term “decision latency.” Decision latency describes that gap between when an anomaly/alert/problem is detected and when an action is taken. The activities in that gap typically look like:
Anomaly detected –> Expert receives anomaly –> Expert assesses validity –> Expert compiles personal experience, relevant data sources, and required tools –> Expert applies domain skills to analyze all available information –> Expert decides on best course of action –> Expert routes actions to the right team –> Action taken
Everything between “Anomaly detected” and “Action taken” is reliant on experts. And current expert decision capacity means this workflow is executed for less than 5% of anomalies detected. With more data and more insights being surfaced than ever before, this traditional human-reliant decision workflow cannot scale.
And this structural gap is only getting worse.
An Expertise Crisis That’s Already Here
It’s not a new conversation, but it’s still a relevant one. The workforce that built deep operational expertise over decades is retiring — and it cannot be replaced fast enough. Studies consistently show 30–40% of the skilled manufacturing workforce will be eligible for retirement this decade. The tacit knowledge held by a veteran reliability engineer or operator is walking out the door faster than any hiring pipeline can compensate.
Organizations that survive this won’t just be the ones that attract talent. They’ll be the ones that institutionalize expertise — encoding the knowledge, experience, and intuition of their best experts into systems that can act on it at machine speed, without routing every decision through a human who may not have bandwidth when the moment demands it. Agentic Operations makes this possible.
Decision Acceleration with Agentic Operations
We measure our platform’s value in decision acceleration: the systematic reduction of time between a condition arising and the right action being taken. Alerts and insights alone don’t generate ROI. Decisions made = outcomes delivered.
LNS Research has put a name to the technology category that will enable autonomous decision intelligence. And I genuinely appreciate that. The taxonomy they’ve created provides manufacturing leaders a language to ask better questions and orient themselves in a space that has long needed clarity.
Agentic Operations is here, the industry is ready, and we’re excited to be on the leading edge of this market evolution.
