(Panel) What Next? How to Advance AI/XR/Digital Twin Projects From Pilot to Production

Moving from a successful pilot to production deployment is where many AI, XR, and digital twin initiatives stall. In this session, industry leaders focus on the operational, technical, and organizational work required to scale across teams, sites, and global enterprises.

They’ll share lessons learned from real-world rollouts and explore what it takes to operationalize these technologies in the workplace, from workflow integration and governance to scaling infrastructure, content, and adoption over time. Hear about:

• Scaling architecture & integration: Standardizing across cloud, edge, and on-prem; integrating with enterprise systems; ensuring interoperability   
• Data pipelines: Ensuring reliable data flow across systems, tools, AI models, simulations, and real-time applications
• Device & platform management: Managing XR, edge, and spatial environments; supporting multiple devices, operating systems, and AI/visualization frameworks
• AI & digital twin operations: Managing model lifecycles and simulation environments (training, versioning, monitoring, physical-digital synchronization)
• Performance & reliability: Addressing latency, bandwidth, compute, and streaming constraints; establishing SLAs, monitoring, telemetry, and support 
• Content, model & simulation management: Scaling creation, versioning, and distribution of XR content, AI models, and simulations with consistent quality
• Security & compliance: Enforcing security policies, identity, and access controls; protecting sensitive data; meeting regulatory requirements
• Cost & governance: Managing total cost of ownership across hardware, compute, licensing, and infrastructure; defining roles, accountability, and decision-making models
• Change management: Driving adoption, embedding new tools and workflows into daily operations, and sustaining training at scale
• Common pitfalls: Fragmentation, unclear ownership, underestimated complexity, lack of data readiness, insufficient planning