Machines Reach for Scale Not Just Interest

Eight out of ten factories run manual, automation beckons. Deployment data shows 98% of manufacturers are exploring or considering AI-driven automation, yet only 20% are prepared to scale it, a gulf that keeps many plants stuck at pilot projects rather than transition moments.

The numbers come from Redwood Software’s Manufacturing AI and Automation Outlook 2026, a snapshot that echoes what many plant managers feel: ambition is high, but the leap to full-scale implementation is the hard part. The case study reports that in the United States, something like 80% of manufacturing facilities have no automation at all. The hardware is willing, but the software remains weak, a gap that frustrates even the most optimistic operations teams. Intrinsic’s public commentary on the space frames the challenge bluntly: robots will not work in most factories if the software and data ecosystems aren’t ready to support them. The sentiment is echoed by practitioners who see value in automation but worry about the integration wrinkles that turn a glossy business case into a stalled project.

Industry voices emphasize a practical path forward. James Taylor, chief commercial officer with OnRobot, has argued that the best entry is to start simple, automate bite-sized tasks, and prove value before expanding. Rick Faulk, chief executive of Locus Robotics, has echoed the sentiment, noting that even smaller plants can automate portions of material handling and repetitive tasks without taking on a full factory automation overhaul. The case study reports a consensus that the easy wins matter, but only if they are designed with scale in mind from the start.

From a CFO and plant-operations lens, the ROI conversation must center on cycle times and throughput as leading indicators of value. Automation should translate into faster cycle times for critical operations and higher throughput where bottlenecks constrain output. Yet the reality remains that many projects stall on integration requirements. Modern automation rarely lives in a vacuum; it must talk to existing controls, data historians, MES and ERP layers, and the plant’s safety and maintenance regimes. The deployment data suggests that without interoperable software architectures and robust data pipelines, the promised gains evaporate as soon as the first production line is wired to a new robot or cobot.

The practical path forward is a disciplined, staged approach that aligns with the operational metric mindset. For managers, that means selecting tasks with the clearest ROI signals, repetitive, low-variance cycles where throughput can be meaningfully increased, or cycle times can be shaved without sacrificing quality. It also means setting clear exit criteria for pilots and designing automation so it can be extended or retracted with minimal disruption. The case study's guidance to “start simple” is not a retreat; it is a strategy to de-risk the broader automation agenda by building confidence, data, and capability before committing to large, multi-line deployments.

Two key practitioner insights emerge. First, hardware readiness is not enough. Without reliable software, data connectivity, and control-system compatibility, automation projects stall after the first week of debugging. Second, integration complexity is the silent killer of ROI. Projects must plan for data migration, integration with PLCs and historians, and the long tail of maintenance and software updates. These realities explain why the 80% figure exists even as nearly all manufacturers say they will automate: interest is high, but execution requires a deliberate architecture, real-time data discipline, and an approach that ties improvements in cycle time and throughput to a transparent financial plan.

In the near term, expect automation to advance in modular, software-driven increments that emphasize interoperability, with vendors competing on how smoothly their stacks slip into existing plant ecosystems. The takeaway for plant leaders is clear: lead with the operational metric, measure cycle times and throughput, and align the automation journey to a credible ROI narrative that can survive the inevitable integration headaches.

Sources