Drones, AI & IOT - What's all the buzz about ?
summary
Shows concrete stacks and results: uses Azure Cognitive Services for OCR and face/emotion JSON, pipelines via Azure IoT Hub and IoT Edge to run models near devices, and deploys models to Jetson TX2 for real‑time drone inference. Gives numbers: a power‑line run detects 217 objects with 6 potential faults over 6 km, Project Brainwave processes ~500 images/sec (US land cover in ~10 minutes for ~$80), and a temp sensor hardware cost spikes to ~$300 leading to $700–800k device budgets. Describes synthetic data via AirSim and a working shark‑detection/drone rescue prototype with life‑buoy drops.
AI-written summary · about
chapters (8)
- 0:00Start
- 1:56Conversational UI Business Scenarios
- 16:40Azure loT reference architecture
- 29:50Azure loT Edge Deployment Cognitive Services Vision
- 31:05Azure MxChip
- 38:13The challenge of small object recognition
- 53:11Streaming drone footage from gaming PC
- 55:51ML processing on drone-direct cloud connection
related talks
33mAI on the Edge – Unlocking New Opportunities in IoT
01h 00m5 Levels of AI in Microsoft Azure
24mWhat does AI mean for IoT development?
50mAzure AI, Power new possibilities for every organization
01h 03mEasier AI allows everyone to build smart apps
24mDon't want to miss out on the AI wave?
27mGenerative AI and IoT, is there a match?
57mEasier AI allows everyone to build smart apps