Autonomous Operations Isn't About Removing People

Overview

Autonomous operations is not primarily about removing people from industrial operations, although workforce implications may certainly follow over time. It is about scaling human expertise across more assets and decisions while acknowledging the current state of technology. While many discussions of autonomous operations focus on lights-out facilities and fully automated decision-making, the reality for most process industries is more practical and more nuanced. The near-term objective is to improve operational performance by reducing routine work, minimizing human exposure to hazards, accelerating decision-making, and enabling scarce experts to support more assets with a broader span of control.

Process manufacturers face an increasingly difficult combination of workforce shortages, growing operational complexity, aging infrastructure, cybersecurity concerns, environmental and safety regulations, energy transition pressures, and demands for greater business agility. Autonomous operations initiatives are emerging as a practical response to these challenges. However, organizations that view autonomy primarily as a workforce reduction strategy risk weakening both the business case and their ability to preserve and scale scarce expert capacity.

This Insight article examines why autonomous operations should be viewed as an expertise amplification strategy rather than a workforce elimination initiative, what has changed to make higher levels of autonomy achievable, and what this means for owner-operators as they develop operating models that apply human judgment across more assets and decisions.

However, organizations that view autonomy primarily as a workforce reduction strategy risk weakening both the business case and their ability to preserve and scale scarce expert capacity.

Autonomous Operations Depends on Human Expertise

Discussions of autonomous operations often evoke images of dark control rooms, unmanned facilities, and plants operating indefinitely without human intervention. These concepts may have value in narrowly defined applications, but they do not accurately reflect the operational realities of most process industries.

Refineries, chemical plants, LNG facilities, pipelines, offshore platforms, mining operations, and power generation facilities operate in highly dynamic environments where business objectives, process conditions, equipment health, market conditions, regulatory requirements, and safety concerns constantly change. These environments require judgment, prioritization, and adaptation that remain heavily dependent on human expertise.

The most successful industrial organizations recognize that expert capacity is becoming increasingly scarce. Many sectors are facing retirement-driven workforce transitions while also struggling to attract and train new personnel with equivalent operational experience. In many organizations, critical operating knowledge still resides in a relatively small number of individuals whose judgment directly influences safety, reliability, production performance, and asset utilization.

This creates a fundamental challenge. The problem is not that organizations have too many people. The problem is that scarce expertise cannot always be allocated to the assets and decisions where it is most needed.

Autonomous operations initiatives should therefore be viewed as mechanisms for scaling expertise, amplifying knowledge, and extending experts’ span of control rather than eliminating personnel.

Rather than assigning experienced personnel to repetitive tasks such as routine surveillance, data collection, alarm response, report generation, or manual coordination activities, autonomous systems can automate many of these activities and reserve scarce expert capacity for higher-value work such as:

• Production optimization

• Exception management

• Strategic decision-making

• Reliability improvement

• Process performance enhancement

• Safety risk mitigation

• Continuous improvement initiatives

The ultimate objective is not fewer people. The objective is better use of human judgment and expertise.

Today's Operations Model Is Becoming Unsustainable

For decades, industrial operations have relied on staffing models built around local decision-making, physical presence, and significant manual intervention. These approaches evolved when information was fragmented, communication capabilities were limited, and many operational decisions had to be made close to the process.

That environment is changing rapidly.

Many organizations now struggle to maintain consistent staffing levels across multiple facilities. Remote facilities are often particularly difficult to support because specialist expertise is concentrated in regional centers or corporate headquarters. At the same time, assets are becoming more interconnected, more instrumented, and increasingly data-intensive, expanding the number of situations that compete for limited expert attention.

Operators and engineers today frequently face a different challenge from that faced by previous generations. Rather than suffering from insufficient information, they are overwhelmed by it.

Large operations routinely generate:

• Millions of process measurements

• Thousands of alarms

• Equipment diagnostic data

• Maintenance records

• Laboratory results

• Business system information

• Environmental and sustainability metrics

However, adding more data does not automatically improve decision-making. In many cases, personnel spend significant time searching for information, validating data, coordinating with multiple departments, and performing routine analysis before expert judgment can be applied.

As industrial organizations pursue higher levels of operational excellence, simply adding more personnel is becoming more difficult and less economically attractive. At the same time, expecting existing specialists to absorb continually increasing complexity without better allocation of their expertise is unrealistic.

This growing mismatch between operational complexity and available expert capacity is one of the primary drivers behind autonomous operations initiatives.

What Has Changed?

The concept of autonomous operations is not new. What has changed is the availability of technologies capable of supporting it at industrial scale.

Several trends have converged in recent years.

Better Operational Visibility

Industrial organizations have invested heavily in connectivity, instrumentation, historians, industrial data platforms, and contextualized information systems. As a result, operational data is more accessible than ever before.

This visibility provides the foundation required for advanced analytics, predictive applications, and decision support that can amplify expert knowledge across a wider operating scope.

Advances in Industrial Artificial Intelligence

Recent developments in artificial intelligence are enabling organizations to move beyond traditional monitoring and alerting systems.

Industrial AI technologies are increasingly capable of:

• Identifying anomalies

• Detecting emerging equipment problems

• Recommending corrective actions

• Prioritizing operator attention

• Optimizing operating conditions

• Supporting complex planning decisions

Importantly, most current industrial AI deployments still function as advisors rather than replacements for human decision-makers.

This distinction is critical. The greatest value today often comes from accelerating and extending expert decision-making rather than fully automating it.

Remote Operations and Centralized Expertise

Many industries have demonstrated the value of centralizing expertise through integrated operations centers and remote operations centers, increasing the span of control available to experienced personnel.

These operating models allow a smaller group of experts to support multiple facilities more consistently. Rather than requiring every site to maintain a full complement of specialists, organizations can allocate scarce expert capacity where and when it is most needed.

Autonomous operations extend this concept further by automating routine monitoring and coordination activities, amplifying reusable knowledge, and escalating exceptions to the appropriate human experts.

Improved Digital Workflows

Modern workflow systems increasingly connect operations, maintenance, engineering, and business functions.

This integration reduces delays associated with manual communication, information handoffs, and decision approvals.

As a result, organizations can execute operational decisions faster while preserving clear accountability and directing human judgment toward decisions that require it.

The Real Destination Is Human-Machine Collaboration

One of the most common misconceptions about autonomous operations is that autonomy increases as human involvement decreases.

In reality, leading autonomous operations programs often increase the value and reach of human expertise while changing how that expertise is allocated.

The future operating model is best understood as a partnership in which autonomous systems extend the span of control of human experts and direct their attention to ambiguity, exceptions, and consequential trade-offs.

Table 1: From Traditional Operations to Autonomous Operations

In this model, technology handles routine, repetitive, and data-intensive activities, while humans focus on ambiguity, judgment, creativity, trade-offs, and strategic thinking. The result is not simply the automation of work, but knowledge amplification across a broader set of assets and decisions.

Organizations should not measure progress toward autonomy by counting how many people are removed from the workflow. They should evaluate how effectively expertise is allocated, amplified, and scaled across the enterprise.

The most mature organizations increasingly measure outcomes such as:

• Faster decision cycles

• Reduced operational risk

• Improved reliability

• Improved energy performance

• Better expert allocation and utilization

• Enhanced safety performance

• Greater organizational agility

These outcomes depend heavily on effective human-machine collaboration and better use of scarce human judgment.

Why the "Lights-Out" Narrative Can Be Harmful

The persistent focus on lights-out operations creates unnecessary resistance in many organizations.

Employees often interpret autonomy initiatives as workforce reduction programs rather than efforts to extend expertise and improve operations. This perception can create skepticism, slow adoption, and reduce engagement.

In reality, many owner-operators already face significant workforce constraints. Experienced operations personnel, reliability specialists, process engineers, and maintenance experts remain difficult to recruit and retain.

The challenge is rarely finding ways to eliminate expertise.

The challenge is finding ways to effectively scale it.

Organizations that position autonomous operations as a workforce reduction exercise may experience:

• Reduced workforce engagement

• Resistance to technology adoption

• Greater organizational change challenges

• Loss of critical institutional knowledge

• Reduced trust in AI-enabled systems

Organizations that position autonomy as an expertise-enablement strategy often achieve better outcomes because employees can see how technology improves safety, reduces frustration, and increases the reach and value of their judgment.

Successful programs consistently communicate that automation should remove low-value tasks, preserve scarce expert capacity, and allow people to concentrate on work requiring judgment.

Implications for Process Industry Operators

As owner-operators develop autonomous operations roadmaps, they should resist the temptation to focus exclusively on technology architectures. The central design question is how expertise will be distributed across assets, roles, and decisions.

Technology is only one component of an operating model designed to scale expertise.

The most successful programs address people, processes, decision rights, operating models, and technology together so that specialist knowledge can be applied consistently where it creates the greatest value.

Organizations should especially focus on the following questions:

• How can scarce expertise be shared across more assets?

• Which activities are repetitive and suitable for automation?

• Which decisions require human judgment?

• How should roles and experts’ span of control evolve as autonomy increases?

• What skills will future operators and engineers require?

• How should centralized and local operations allocate expertise and decision responsibility?

• How will accountability be maintained in autonomous environments?

The answers to these questions often determine success more than the selection of any individual technology platform.

Autonomous operations should ultimately be viewed as an operating model for scaling expertise, not simply a technology deployment.

Recommendations

ARC believes the road to autonomous operations should begin with a realistic understanding of the destination. For most process industry organizations, the objective is not fully unmanned operations. The objective is an operating model that combines human expertise and intelligent automation to scale knowledge, improve expert allocation, and achieve higher levels of safety, performance, reliability, and agility.

Based on ARC research and analysis, we recommend the following actions for owner-operators and other technology users:

1. Anchor autonomous operations initiatives in clearly defined business outcomes such as safety, reliability, responsiveness, energy performance, workforce effectiveness, and organizational agility rather than pursuing autonomy as an abstract technology goal.

2. Reframe autonomous operations as an expertise amplification initiative rather than a workforce reduction initiative.

3. Focus initial efforts on automating routine and repetitive activities that consume scarce expert capacity.

4. Develop operating models that allow scarce expertise to support multiple facilities through centralized and remote operations concepts and a broader span of control for experts.

5. Invest equally in workforce development, knowledge amplification, role design, and technology deployment.

6. Measure success using operational outcomes such as reliability, safety, responsiveness, and workforce effectiveness rather than headcount reduction.

7. Establish clear decision boundaries defining which actions can be automated and which require human judgment or oversight.

8. Create a long-term roadmap that progressively expands autonomy while preserving accountability, safety, operational resilience, and access to expert judgment.

The organizations most likely to succeed on the road to autonomous operations will not be those that remove the most people from operations. They will be those that most effectively combine human expertise with intelligent automation, scale scarce expert capacity across assets, and reserve human judgment for the decisions where it creates the greatest value. Autonomous operations is ultimately about amplifying human capability, not replacing it. If autonomous operations is not about removing people, then what distinguishes autonomy from automation? The answer is not technological sophistication alone. It is who, or what, has authority to act.

For further information or to provide feedback on this article, please contact your account manager or the author at [email protected]. ARC Insights are published and copyrighted by ARC Advisory Group. The information is proprietary to ARC, and no part may be reproduced without prior permission from ARC.