Automation adoption is surging at trade shows. Most US factories still haven't started.

More than 50,000 people registered for Automate 2026, making it the best-attended edition in the event's 10-year history, according to Robotics Tomorrow. The show featured 1,230 exhibitors, with humanoid robots and industrial AI commanding the most floor traffic. The energy was undeniable. The adoption numbers back home are a different story.

Intrinsic CTO Brian Gerkey recently cited a figure that should give every operations leader pause: 80% of US manufacturing facilities currently run zero automation, as reported by Manufacturing Dive. That gap between conference enthusiasm and shop-floor reality is one of the defining tensions in industrial technology right now.

Belief outpaces deployment

The belief-to-deployment disconnect is well-documented. Deloitte's 2025 Smart Manufacturing and Operations Survey found that 92% of manufacturers consider smart manufacturing the primary driver of competitiveness over the next three years, according to Manufacturing Dive's reporting by Sakshi Udavant. Widespread deployment, however, remains the exception rather than the rule.

Jeff Burnstein, president of the Association for Advancing Automation, told Manufacturing Dive that interest is high across the board, but execution is where things get difficult. Cost, integration complexity, and workforce readiness all slow the path from pilot to production scale. Those barriers are consistent regardless of company size, though smaller manufacturers face them most acutely.

The contrast with automation leaders in Asia, where fully lights-out facilities have become a benchmark, adds competitive urgency for US procurement and operations teams evaluating when to move.

Vendors are moving to lower the entry bar

Several recent commercial moves suggest suppliers have heard the message and are actively working to reduce friction. ABB Robotics and Roche announced a collaboration to deploy physical AI in laboratory settings, with early applications targeting pathology slide handling and autonomous connectivity in Core Lab intralogistics, per Robotics Tomorrow. The deployment will combine autonomous mobile manipulation with fixed articulating robots, a configuration that could serve as a reference model for regulated-environment automation beyond pharma.

On the research and development side, InDro Robotics launched a platform called Axiom, described by Robotics Tomorrow as a budget-friendly, humanoid-style robot aimed at physical AI research. The company reported that first customer units are already shipping, signaling that the humanoid category is beginning to reach cost points accessible outside large-scale OEM programs.

Meanwhile, ENCY Software and Stäubli Robotics signed a global agreement to integrate ENCY's CAD/CAM-based offline programming and simulation tools with Stäubli's industrial robots, according to Robotics Tomorrow. Offline programming reduces the time technicians spend on the physical floor during deployment and reprogramming, which directly addresses one of the recurring cost objections manufacturers cite.

Physical AI moves into specialized environments

The ABB-Roche deal is one of the more concrete examples of physical AI, a term describing robots that perceive and act in unstructured environments, moving from warehouse and factory floors into highly controlled settings like clinical labs. Pathology slide handling requires precision and repeatability that autonomous systems are only now reliably achieving at commercial scale.

Orbbec also surfaced at the same moment with a robot-free data collection hardware platform designed to capture real-world demonstrations for physical AI training at scale, per Robotics Tomorrow. The device category matters because data scarcity is one of the primary bottlenecks slowing physical AI model development; tools that let operators record task demonstrations without a deployed robot on-site compress the development timeline.

What this means for your team

• Audit your facility's automation baseline before the next budget cycle. With 80% of US plants at zero automation, the competitive window for early movers remains open, but the technology is maturing fast.
• Evaluate offline programming and digital-twin tools as a first step. Platforms like ENCY Robot reduce deployment risk and floor disruption without requiring full system integration upfront.
• Watch the humanoid and physical AI category for cost inflection. InDro's Axiom and similar entrants suggest price points are dropping; set a formal review trigger tied to cost-per-task benchmarks relevant to your operation.
• For regulated environments such as pharma, food, or clinical labs, the ABB-Roche collaboration provides a reference architecture worth requesting details on from your automation integrator.