filmov
tv
IoT Live DevNation 2015

Показать описание
Tuesday, June 23rd
We provided Bluetooth Low Energy beacons to 300 of the early DevNation attendees, collection over 7 million sensor readings from those individuals as the navigated around the DevNation event. Our sensors were based on the Raspberry Pi. Around 8:00 AM, you see most folks heading up to the 3rd floor for breakfast and then hitting various breakout sessions and about 4:00 PM (16:00), you see the audience arriving in time for the keynote - where they were able to see all of their aggregated movements on this map. After the keynote, you see people trickling out of the event and heading off for their various evening activities.
Our data capture was imperfect as this was our first POC of using BLE for tracking but it makes for a fantastic visual replay of the actual events - all via a HTML5, web-based dashboard.
This map was our real-time dashboard allowing us to watch traffic patterns but it also offers a playback feature to rip back through N hours of the Apache ActiveMQ + Spark processed datastream.
Key technologies include:
Bluetooth Low Energy beacon
Raspberry Pi - based BLE sensors
JBoss Fuse A-MQ for the massive data ingestion
Apache Spark Streaming for the real-time analysis (what room is the beacon in)
We provided Bluetooth Low Energy beacons to 300 of the early DevNation attendees, collection over 7 million sensor readings from those individuals as the navigated around the DevNation event. Our sensors were based on the Raspberry Pi. Around 8:00 AM, you see most folks heading up to the 3rd floor for breakfast and then hitting various breakout sessions and about 4:00 PM (16:00), you see the audience arriving in time for the keynote - where they were able to see all of their aggregated movements on this map. After the keynote, you see people trickling out of the event and heading off for their various evening activities.
Our data capture was imperfect as this was our first POC of using BLE for tracking but it makes for a fantastic visual replay of the actual events - all via a HTML5, web-based dashboard.
This map was our real-time dashboard allowing us to watch traffic patterns but it also offers a playback feature to rip back through N hours of the Apache ActiveMQ + Spark processed datastream.
Key technologies include:
Bluetooth Low Energy beacon
Raspberry Pi - based BLE sensors
JBoss Fuse A-MQ for the massive data ingestion
Apache Spark Streaming for the real-time analysis (what room is the beacon in)