Roadmap to physically intelligent industrial operations

Editor’s note: This is the fourth and final part of our series, the Smart Operations Playbook, brought to you by four SMEs at ARC Advisory Group.

In Part 1, Colin Masson made the case for "taming the data beast" by establishing industrial data fabrics.

With Part 2, ARC's Craig Resnick and Interpreet Shoker filled a very important role in our coverage of AI's impact on the manufacturing workforce by emphasizing how artificial intelligence should augment industrial roles—not replace human employees.

And in Part 3, Patrick Arnold wrote on how intelligent robots are bridging the gap from automation to autonomy.

We’ve covered significant ground in this series.

We began our journey by tackling the single biggest barrier to digital transformation: data chaos. We made the case that “taming this beast” with an industrial data fabric is the unskippable first step.

From there, we built on that foundation. My colleagues Craig Resnick and Inderpreet Shoker showed how this clean, contextualized data is the fuel for practical industrial AI and connected worker platforms, framing AI not as a job-replacer but as a powerful tool to augment and empower your workforce.

Most recently, Patrick Arnold connected that intelligence to the physical world, exploring the new era of Physical Intelligence and how edge computing is enabling smart, secure, and functional robotics.

See also: Adoption of automation, AI-powered tools accelerating across sectors, survey shows

Now, in this final installment, we must tie it all together. But as we do, we must acknowledge a stark reality revealed in our latest research: the gap between the leaders and everyone else isn't just widening; it’s fracturing.

The Great Divergence: Insights from the Q4 pacesetter survey

ARC Advisory Group will do a deeper dive into its Q4 2025 Industrial AI Pacesetter Survey during ARC’s upcoming Industry Leadership Forum in February, but the survey included 510 industrial respondents, reveals what I call “The Great Divergence.”

The market has bifurcated into distinct species: the pacesetters, or leaders, made up of about 12.9% of the market; the mainstream (55.3%); and the laggards (31.8%).

The data indicates that pacesetters are operating on a fundamentally different strategic plane. While laggards remain trapped in “pilot purgatory,” obsessing over incremental cost reduction and treating AI as a siloed IT project, pacesetters have weaponized industrial AI as a core business strategy.

Crucially, their motivations differ. Pacesetters cite “addressing skills gaps and augmenting the workforce" as their primary driver. Laggards, conversely, are still focused almost exclusively on “reducing costs.”

See also: How smart industry is escaping the 'single cloud of failure'

Pacesetters understand that in an era of labor shortages, the goal is to create a compound advantage through three strategic pillars: structural (data fabrics), human (the connected worker), and physical (physical intelligence at the edge).

Step 1: Assess your maturity (where are you?)

The first step in any journey is knowing your starting point. Based on our survey data, you likely fall into one of these categories:

• Laggards (The 31.8%): You are at risk of irrelevance. Your data is trapped in silos, and you have “islands of automation” with no single source of truth. My advice: Stop buying AI. You aren't ready. Focus 100% of your budget on basic connectivity and getting data out of stranded machines.
• Mainstream (The 55.3%): You have successful pilots but are stuck in “pilot purgatory.” You lack full IT/OT/ET convergence. My advice: Declare a moratorium on “science experiments” and redirect that budget into a data fabric infrastructure.
• Pacesetters (The 12.9%): AI is embedded in your strategy. You are seeing measurable ROI. My advice: Scale the edge and pivot toward agentic AI to further empower your workforce.

Step 2: Find the value (where do you start?)

Do not start with the "newest" technology. Start with your most valuable problem:

• High-impact downtime: Use your data fabric to build your first predictive maintenance model for the one asset that causes the biggest cascade of failures.
• Quality escapes: Deploy AI-powered machine vision where you produce the most scrap.
• Ergonomic nightmares: Find the task causing the most worker injuries; that is your first use case for a collaborative robot.

Step 3: Build the team (who will do it?)

Technology alone solves nothing. The single greatest differentiator between leaders and laggards is people. You must break down the walls between IT, OT, and engineering. This is a cultural change, not just an organizational one.

See also: The strategic importance of industrial data fabrics

You need “translators”—professionals who speak the languages of both operational problems and technical solutions. Pacesetters are already cultivating what we call the “synapse worker,” an empowered professional who collaborates with AI to drive new levels of value.

A phased rollout: Your playbook

Here is your final, phased playbook for building a smart operation:

• Phase 1: Foundation (The first 12 months): Launch your industrial data fabric initiative. Stop buying software that locks data in proprietary databases; demand open architecture.
• Phase 2: Augmentation (months 12-24): Deploy high-ROI, practical AI solutions on top of that data foundation, such as predictive maintenance or connected worker platforms to preserve expertise.
• Phase 3: Action (years 2-4): Begin deploying physical intelligence. Translate digital insights into physical action with fleets of AMRs (autonomous mobile robots) or cobot workcells that adapt to their environment.

This is not a one-time project; it is a continuous journey from reactive chaos to predictive, physically intelligent operations.

By following this playbook, you move beyond the "AI wars" hype and toward a more resilient, efficient, and human-centric industrial future. That is the very definition of a smart operation.