Sight Machine introduced autonomous AI agent crews for manufacturing at Hannover Messe 2026, deploying coordinated teams of specialized AI agents that continuously monitor and optimize production performance. The agents work on top of Sight Machine’s semantic data layer and communicate via Model Context Protocol (MCP), according to AIThority.
The system addresses a gap that general-purpose AI tools cannot fill: raw industrial data requires domain-specific context to be useful. Sight Machine’s semantic layer translates machine-level sensor data into structured representations of manufacturing processes. The agent crews operate on that structured data rather than trying to interpret raw signals directly.
How the Agent Crews Work
Individual agents focus on specific KPIs: throughput, quality, defect rates, cost. They coordinate as a crew to balance tradeoffs, such as maximizing throughput versus maintaining quality standards when switching product types or raw materials.
“The agents are always on, looking at current operating conditions, looking at all the process parameters, looking for anomalies and changes in conditions, and providing recommendations to operators on how they can improve,” Sight Machine VP of Product Andrew Home told AIThority.
The crews use Sight Machine’s manufacturing analytics tools: statistical analysis, machine learning techniques like random forest models, centerlining, golden runs, and root cause analytics. Within a permissioned framework, agents can also code and orchestrate new tools for analysis and presentation.
The deployment model is progressive. Agent crews initially operate in analysis-and-recommendation mode, simulating possibilities and presenting options to operations teams. As the crew demonstrates reliability, manufacturers can extend agent authority to directly control specific settings or process steps. The pace of that transition is set by the manufacturer, not the software.
Distribution Through Existing Tools
Sight Machine is making agent outputs accessible through Microsoft Excel, Microsoft Teams, Microsoft Fabric, and Databricks, rather than requiring operators to learn new manufacturing software. The agents themselves run on Microsoft Azure with the platform’s existing corporate security and data governance controls.
CEO and co-founder Jon Sobel told AIThority: “Sight Machine has bridged the gap between general-purpose AI technology and the extremely challenging nature of production data.”
The capabilities are being previewed at Hannover Messe with expanded availability planned for later in 2026.