Connected Worker Manufacturing 2026: From Industrial Data to Action with Agentic AI

How can manufacturers turn growing interest in AI into practical improvements on the plant floor?

At the Connected Worker Manufacturing Summit, held October 13–15 in Chicago, KMS Technology joined SAS to explore that question with manufacturing, operations, and digital transformation leaders. Our conversations highlighted several priorities shaping their next steps:

• Building data teams can trust. Leaders described operations still relying on paper, spreadsheets, and disconnected legacy systems. Capturing and normalizing operational data remains essential to understanding performance, including cycle times, and identifying where to improve.
• Creating consistency across plants. Differences in how sites collect data, run processes, and measure quality make it difficult to compare performance. Leaders are looking for clearer visibility across locations to identify gaps and share effective practices.
• Connecting fragmented tools and platforms. Conversations surfaced the challenge of managing different vendors across functions and piecing together individual AI tools. There was interest in a more integrated view of safety, maintenance, quality, and operations.
• Helping frontline teams work safely and effectively. Worker safety, training, and engagement were recurring topics, alongside concerns about a retiring workforce. The opportunity is to make operational knowledge and timely guidance more accessible to the people doing the work.
• Starting with a focused operational need. Priorities ranged from worker safety and quality to predictive maintenance and parts management. Some organizations were exploring an initial use case; others were looking to extend existing analytics into more coordinated action.

These conversations reinforced the need for an adoption path that reflects each manufacturer’s operational priorities and data maturity.

That was the focus of our joint workshop, “Building a Safer Future with the Agentic Workforce,” where we explored how agentic AI can connect trusted data, operational context, and enterprise tools to shorten the journey from detecting an issue to taking appropriate action.

Closing the gap between insight and action

Sensors, cameras, and analytics can identify equipment anomalies or unsafe conditions. But responding often still requires someone to find the right procedure, review maintenance history, check available resources, and weigh the impact on production.

Our discussion focused on how agentic AI can bring those steps together. By connecting operational data with business rules, documentation, and enterprise tools, agents can help teams evaluate a situation, recommend corrective action, and initiate approved workflows.

An illustrative maintenance scenario made the distinction tangible: a general AI assistant might explain common causes of rising pump vibration. An agent connected to plant systems could combine a failure prediction with maintenance procedures, parts availability, and technician schedules to draft a work order for human approval.

Three practical paths to adoption

We explored three ways manufacturers can put these capabilities to work:

• Make faster decisions. Combine sensor alerts with equipment manuals, maintenance records, and work instructions to help teams understand an issue and identify an appropriate response. The industrial fan example illustrated how anomaly detection and contextual information can support maintenance decisions.
• Build analytical capabilities. Use natural-language instructions with SAS Viya tools to support data exploration, feature engineering, predictive modeling, and evaluation. The predictive maintenance example walked through identifying equipment at risk of failure and assessing the tradeoffs between alerts and missed failures.
• Adopt an integrated solution. The Industrial Command Center approach brings together SAS AI and analytics, Lenovo edge hardware, and KMS engineering to support worker safety, predictive maintenance, and broader operational use cases using existing cameras, sensors, and systems.

Building trust into operational AI

Faster decisions must also be dependable. We discussed the foundations required for responsible adoption: trusted data, explainable recommendations, secure system access, clear autonomy limits, and human oversight of critical decisions.

Ownership, monitoring, and ongoing evaluation matter just as much after deployment. Manufacturers need to know what an agent can do, when it must escalate, and who remains accountable for the outcome.

Together, SAS and KMS bring complementary capabilities to this challenge: an AI and analytics platform paired with the engineering, integration, deployment, and ongoing support needed to apply it in industrial environments.

Thank you to everyone who connected with us during the summit and at Booth 38. We look forward to continuing the conversation on turning industrial intelligence into safer, more effective operations.