A semantic Internet of Things
summary
Prototype claims the Internet of Things becomes semantic by turning raw sensor data into human‑style facts (a bank charge → romantic dinner), using Google Cloud tools and graph storage. Implementation glues Cloud Vision API for image labels, a local microcomputer for edge preprocessing, and a Dgraph graph database plus decentralized ontologies mapped via schema.org RDF‑IoT to store relations. Example: processing 2,500 ship cameras at five‑minute intervals. Prototype runs but faces misclassification and privacy challenges; handling false relations remains unresolved.
AI-written summary · about
chapters (10)
- 0:00Introduction and background
- 3:54Breaking the digital-physical wall
- 6:47Digital vs physical world disruption
- 12:06The database problem in IoT
- 17:10Human observation and language
- 21:44Semantic Internet of Things vision
- 25:13Mapping physical world to knowledge
- 27:00Superposition project overview
- 31:15Future of semantic IoT
- 31:57Q&A: Privacy and data handling
related talks
31mDon't be Religious about the Internet of Things!
01h 03mPolicy implications of the internet of things
22mGeneric and Extensible web of things manager using JSON schema & AI.
43mIndustrial IoT in practice
21mKeynote: Can the Web Save the Internet of Things?
50mThe Internet of Natural Things
05mIoT Inspector: a platform for real-world smart home research
09mInterview – Yodit Stanton