General Keynote Multi-tiered digital twins with what-if analysis...
summary
A Microsoft-built multi‑tier digital twin ingests systems‑of‑record, IoT telemetry and partner‑shared data into a graph and runs what‑if analysis plus automated prescriptions. Predictive and prescriptive AI (OR plus ML, simulated annealing) lift a baseline fitness from ~85 to ~97 and cut units on hand from ~1,000 to ~880 under normal settings. When SLA tightens to 5 days the policy flips to ~90% make‑to‑stock and on‑hand rises to ~1,422 (~40% increase) while still beating the baseline by ~10 points; graph queries expose single‑port bottlenecks and simulate node/edge failures.
AI-written summary · about
chapters (10)
- 0:00Intro
- 0:18Agenda
- 1:04Supply chains
- 3:53Complex systems
- 5:07Modeling framework
- 7:24Digital twins
- 12:20Digital twins examples
- 15:02Graph right
- 16:42Analytics and AI
- 19:22AI Framework
related talks
10mAI-Powered Digital Twins: Patterns for Smart Manufacturing
27mBuilding Reliable Human-AI Warehouse Operations at Scale
59mInnovating power plant operations through digital twins
24mDigital Battery Twin
15mStabilizing Ops & Eliminating Variability on
14mTransforming Manufacturing: AI-Driven Predictive Maintenance
01h 23mReal-Time Structural Health Monitoring Using Physics-Driven Digital Twins
13mBuilding Scalable AI/ML Platforms for Industrial IoT