The Evolution and Innovation of Edge Computing in Baidu
summary
Baidu ships a commercial edge platform BIE and open‑sources core tech as Baetyl, with a Docker‑based 1.0 and a June 2020 Beetle 2.0 Kubernetes‑based edge clusters model that lets each edge run as an independent k3s or full Kubernetes cluster and connect via cluster‑to‑cluster orchestration. It adds automatic model conversion to OpenVINO/Movidius for on‑device AI, drives real deployments (dozens of customers for safety and environmental monitoring), model deploy time cut from 24h to 1m, and a V2X setup that lowers autonomous‑driving cost by 60%.
AI-written summary · about
chapters (16)
- 0:00Intro
- 0:28Computing is everywhere
- 1:17Fusion between the Cloud and the Edge
- 2:08Baidu's edge computing system
- 3:13Baetyl, a common Edge computing framework
- 4:04Baetyl 1.0, container centric
- 5:18Baetyl 2.0, cloud-native centric
- 6:41Open ecosystem, Vendor neutral
- 8:44Baetyl and Baidu Al integration
- 9:33BIE Compute Vision Solution
- 10:32Baetyl & Edge Foundry
- 11:23Smart Grid
- 12:25Smart Quality Inspection
- 13:28V2X Autonomous Driving
- 14:23Baetyl 3.0
- 14:45Join us
related talks
05mEdge AI: Because the Cloud is Too Slow...
26mAI computing power in Edge Industry use cases
19mDistributed AI for Automated and Scalable Quality Inspection using EDGE Computing...
14mEdge AI: Real Solutions for Real World Problems
15mSimplifying & Scaling Edge AI
36mWhat's happening in AI
18mThe Cutting Edge of AI and IoT
13mHow Edge AI Will Save The World