Beehive monitoring as an application of Fog Computing

Voxxed Days Belgrade 2016

Transcript

1. Beehive monitoring as an application of Fog Computing
2. Klemen Lilija klemen@relayr.io
3. AGENDA • IoT • Bee monitoring project • Fog Computing
4. INTERNET OF THINGS
5. A global infrastructure for the Information Society, enabling advanced services by interconnecting (physical and virtual) things based on, existing and evolving, interoperable information and communication technologies. - International Telecommunication Union
6. Physical Devices | Networking | Data Aggregation | Applications
7. User Application Sensors & Actuators Cloud Gateway Adapters Service Integration
8. BEE POPULATION MONITORING
9. Bee population decline • bees pollinate around 70% of world’s crops • decrease from 6 million (1945) to 3 million (2005) colonies • colony collapse disorder • possible reasons: • parasites • pesticides • malnutrition • agriculture intensification
10. Bee population monitoring • project originally started by CISCO as CSR • part of The Global Initiative for Honey bee Health (GIHH) • CSIRO, federal government agency for scientific research in Australia
11. APPROACHES
12. CSIRO: RFID tags • 2,5 x 2,5 mm RFID sensor • checkpoints • 1000s of bees in Tasmania already tagged - csiro.au
13. relayr & Cisco: Beehive monitoring - zajtam, flickr Temperature Humidity Audio Weight Image
14. Beehive monitoring: first steps - dclohman, twitter • ESP8266 • WiFi router • relayr Cloud • Sensors: • Temperature • Humidity • Noise level • Weight
15. Bees with RFIDs
16. Easy for beekeepers to use?
17. unreliable networking limited processing power low power Constrained Devices https://tools.ietf.org/html/rfc7228
18. Beehive monitoring: next iteration • WiFi -> 802.15.4 / 6LoWPAN • MQTT -> CoAP • gateway with a Wide Area Network (LoRa) • push ‘smartness’ to the edge of the network
19. 6LoWPAN • IPv6 over Low power Wireless Personal Area Networks • adaptation layer for IPv6 IEEE 802.15.4 based networks • defines encapsulation and header compression mechanism • Zolertia (Contiki) • phyWave (RIOT) - zolertia.io - phytec.de
20. LoRa • long range / good building penetration (up to 15 km) • low energy usage for endpoints • ability to run in unlicensed bands (868/915/433 MHz) • high density of nodes
21. LoRa • LoRa - physical layer (Semtech) • LoRaWAN - defines the MAC layer • The Things Network
22. Beehive monitoring: next iteration • WiFi -> 6LoWPAN • MQTT -> CoAP • gateway with Wide Area Network (LoRaWAN) • push ‘smartness’ to the edge of the network
23. Sensors and actuators. Interface to the global network. Computation and application enabler. Mote Gateway Cloud
24. ESP8266, send data every 10s. WiFi router, forward data packets. AWS, data aggregation, analytics, applications, etc. Mote Gateway Cloud
25. Zolertia, send filtered data. Dell 5100 EDGE, streaming analytics. AWS, further data manipulation & application enabler. Mote Gateway Cloud
26. FOG COMPUTING
27. Definition “Fog computing is a highly virtualized platform, which provides computation, storage, and networking services between the end nodes and traditional clouds.” - Aazam et al., IEEE, 2015
28. Local computing • rack full of servers • Specialized gateway (DELL Edge Gateway 5100) • System on a chip (Raspberry Pi)
29. COMPONENTS
30. Protocol translation • pluggable architecture, with adapters for different protocols • support legacy protocols • lower computing demands for attached devices • low power demands for the attached devices • lower cost of attached devices
31. Sensor data translation • translation of sensor data into a uniform format • The IEEE 1451.4 Standard for Smart Transducers • additional metadata (Web Thing Model, W3C)
32. Data storage • long term data storage • temporary data storage enabling local logic and data preprocessing • database replication • database clustering
33. Local logic • control loop for rapid response industrial systems • low latency • high availability and reliability • rule engine • streaming analytics • local audio and video streaming
34. Transport layer • transport of preprocessed data for service consumption • Encryption / decryption, authentication, privacy • transport abstraction for connection to the data center (HTTP, websockets, raw sockets, CoAP, etc.)
35. Security • reduced attack surface • easier to keeping up with the evolving security protection • attack surface limited to the gateway-thing • mitigate the vastly different expected lifetimes of IT and ‘things’
36. Device Management • registry of all devices that interact with a gateway • devices bootstrapped, added, updated and removed locally • firmware update of the gateway remotely
37. Microservice based
38. WiFi 802.15.4 BLE RS-232 Ethernet Local storage Sensor data translation Local logic Transport layer
39. Core service Cloud Fog Things
40. Cloud Fog Things Core service
41. THANK YOU!