Flexible Manufacturing with AWS and SoftServe: How Simulation-First Robotics Reaches Production Faster

This article describes how SoftServe and AWS demonstrated a simulation-first approach to flexible robotic manufacturing at Hannover Messe 2026, using AI and cloud services to enable adaptive automation that responds to changing customer requests without hard-coded workflows.

• Production cell accepted custom coaster orders via kiosk, coordinated mobile robots, robotic arms, laser engraver, and quality inspection across five-day live demonstration
• Amazon Bedrock Agents orchestrated workflow steps; Nova Canvas generated custom designs from text prompts in 3-4 seconds
• AWS IoT Core and IoT Greengrass enabled bidirectional device communication; SQS and Step Functions managed job queuing and multi-step workflows
• AI-powered vision inspection verified engraving quality automatically using Amazon Nova VLM
• Achieved 45-50 second end-to-end cycle time per coaster with near-100% pick success rate across all five days
• Simulation-first pattern uses NVIDIA Isaac Sim for pre-deployment validation on EC2 GPU instances before physical commissioning

The demonstration proves that flexible automation becomes achievable when AI, cloud services, simulation, and robotics work together, enabling manufacturers to adapt operations as business needs evolve.