Open, Intelligent, and Secure: The Next Industrial Edge
summary
Claims: industrial deployments use 10Base-T1S multi-drop Single Pair Ethernet to get deterministic 10 Mbps links, cutting gateways and cabling. A field rollout uses Arduino Opta hardware with zero issues reported while a maintenance firm prototypes fleet telemetry plus AI for actionable alerts. Edge AI tooling is being packaged into Foundries/Edge Impulse workflows and FPGA toolchains are simplified so hobbyist-grade entry exists — even a $130 PolarFire dev kit for accelerated FPGA inference. Security and 10–20 year fleet management remain concrete constraints.
AI-written summary · about
chapters (8)
- 0:00Industrial edge introduction
- 3:15Open source early adoption
- 7:57IoT security challenges
- 10:53Single pair Ethernet connectivity
- 15:52Maintenance as software
- 17:43AI for predictive maintenance
- 19:53FPGAs for AI development
- 22:46Future of industrial edge
related talks
20mO11y in the Evolution of AI-Era Networking Infrastructure
11mKeynote: LF Edge: Open Source to Enable 5G, IoT and AI Development
11mKeynote: LF Edge: Open Source to Enable 5G, IoT and AI Development
26mAI computing power in Edge Industry use cases
18mThe Cutting Edge of AI and IoT
05mEdge AI: Because the Cloud is Too Slow...
37mHardware At The Edge Reliability, Ai Infrastructure And Defensive Design
10mApproach to AI at a large scale