Most Plants Still Don’t Automate

Eight in ten factories run without automation, even as robots wait on the shop floor.

That disconnect sits at the center of a broad push toward smarter plants and a stubborn reality: the hardware is ready, the software isn’t always. A recent snapshot from Redwood Software’s Manufacturing AI and Automation Outlook 2026 shows a feverish interest in automation, with 98% of manufacturers exploring or considering AI driven automation. Yet only 20% say they’re prepared to scale what they start. In the United States, Brian Gerkey, chief technology officer of Intrinsic, says bluntly that the vast majority of facilities have the hardware ready but the software is not. The hardware is willing, and the software is weak, he told The Neuron Daily.

The numbers aren’t just about early stage pilots. They reflect a broader caution about ROI and readiness. The case is not for lack of appetite; it is about the path from pilots to production. The same outlook notes that nearly all manufacturers are considering automation, but the jump to full scale deployment remains a bottleneck. Deployment data shows that companies that start with small, well defined automation wins tend to reap payback more quickly than those chasing a grand, end to end rollout. The case study reports that the most credible gains come from tasks with clear, repeatable cycles and predictable throughput, not from grand overhauls of entire lines.

Industry practitioners emphasize a pragmatic path forward: start simple and scale. Rick Faulk, CEO of Locus Robotics, and James Taylor, chief commercial officer with OnRobot, both argue that even smaller plants can automate some tasks and build a foundation for broader automation. The guidance is consistent with a hard won reality in many facilities: the ROI hinges on how quickly you can reduce cycle times and increase throughput for targeted tasks, then prove the gains to unlock capital for broader work. In other words, the ROI comes from measurable improvements in the clock, not just software buzzwords.

The practical challenge is integration. Automation that runs on a bench in a lab rarely survives the production floor, unless it can talk to the plant's control systems, data streams, and scheduling software. Integration requirements, connecting new robots, cobots, or vision systems to PLCs, MES, or ERP, and ensuring data security and consistency, often determine whether a pilot becomes a production line. The reality is that many facilities must invest not only in hardware but in the software backbone, data architectures, and operator training to realize meaningful throughput gains. In the words of deployment data and industry veterans, the digital layer is often the harder part of the equation.

That is where skilled trades come into play, not as a side show but as a critical factor in the ROI equation. Automation projects that reach scale typically require electrical and control system expertise to wire, commission, and maintain interfaces with legacy equipment. In many cases, automation augments existing craft labor by taking over repetitive, dangerous, or precision based tasks, while technicians and inspectors adapt to new roles in a more data driven, monitored environment. The exact mix depends on the site, but the pattern is clear: software maturity and integration readiness determine whether robots merely sit idle or actually deliver faster cycle times and higher throughput.

Looking ahead, the industry is not abandoning automation; it is insisting on disciplined, ROI driven execution. The brightest opportunities lie in modular, high utility tasks that can be automated quickly and scaled methodically, with clear data to back every dollar spent. As practitioners point out, the goal is not plug and play miracles but dependable, measurable improvements in operations, with a clear bridge from pilot to production.

The journey remains real, not mythical. The numbers reflect a gap between interest and execution, a gap that savvy manufacturers are choosing to close one task at a time.