Using the FLaNK Stack for edge ai (apache mxnet, apache flink, apache nifi, apache kafka, apache kudu)
3 likes690 views
ApacheCon @Home took place virtually from September 29 to October 2, 2020, featuring discussions on topics such as AI, streaming, and machine learning. Key sessions included incrementally streaming RDBMS data and Apache Deep Learning applications on edge devices. The event showcased the capabilities of the Apache Flink stack and various sensors for data processing and analytics.
Using the FLaNK Stack for edge ai (apache mxnet, apache flink, apache nifi, apache kafka, apache kudu)
- 1. APACHECON @HOME Sept. 29th – Oct. 2nd 2020
- 2. 2 Join Me Virtually @ Future of Data - Princeton on Youtube & Zoom @PaasDev https://www.meetup.com/futureofdata-princeton/ https://github.com/tspannhw From Big Data to AI to Streaming to Containers to Cloud to Analytics to Cloud Storage to Fast Data to Machine Learning to Microservices to ...
- 4. My Talk List 29/9/2020, 1:35 PM EDT - Incrementally Streaming RDBMS Data to Your Data Lake Automagically with John Kuchmek 29/9/2020, 2:15 PM EDT - Using the Mm FLaNK Stack for Edge AI 29/09/2020, 3:35 PM EDT - Real-Time Stock Processing With Apache NiFi, Apache Flink and Apache Kafka with Pierre Villard 30/9/2020, 1:35 PM EDT - Apache Deep Learning 301 with Ian Brooks, PhD. 30/9/2020, 2:15 PM EDT - Edge to AI: Analytics from Edge to Cloud with Efficient Movement of Machine Data with Paul Vidal 30/9/2020, 2:55 PM EDT - Utilizing Apache NiFi and MiNiFi for EdgeAI IoT at Scale with Sunile Manjee
- 5. • • • • • • Edge AI with FLaNK
- 6. APACHECON NA 2020 The FLaNK stack is a team of Apache Open Source software that can be used to rapidly build streaming event pipelines that span use cases from Cloud Ingest, I(I)oT, Log Processing, Cybersecurity Analytics and more. The FLaNK stack consists of Apache Flink, Apache NiFi and Apache Kafka at it’s core. With Apache Kudu as the second option for the K and a preferred data store.
- 8. APACHECON NA 2020 Apache MXNet and DJL.AI as a preferred option for running Deep Learning as part of data flows. We can run Deep Learning on edge devices, in NiFi flows, in Flink apps and in NiFi connected cloud machine learning services like Cloudera Machine Learning.
- 9. 9 Edge AI to Cloud Streaming Pipeline Device Data SensorsEnergy Logs Weather Sensors Aggregates Energy SQL Analytics MiNiFi Agent Deep Learning Classification Edge Private Cloud Multi-Public Cloud
- 11. APACHECON NA 2020 Edge Devices NVIDIA Jetson Xavier NX NVIDIA Jetson Nano RPI 4 NCC2 Enviro+ USB Web Cameras Enviro+ Sensors https://www.datainmotion.dev/2020/06/unboxing-most-amazing-edge-ai-device.html Edge Software Apache NiFi / MiNiFi Java and C++ Agents Python 3 DetectNet GoogleNet MQTT HTTP JDK 8
- 12. APACHECON NA 2020 MiNiFi Java Agent ● Reads Sensor Logs ● OpenVino NCC2 AI ● Reads Images ● Sends to NiFi Gateway Apache NiFi Gateway processors, validates, transforms, cleans, routes and streams events for additional processing through Apache Kafka topics.
- 13. APACHECON NA 2020 {"uuid": "rpi4_uuid_jfx_20200826203733", "amplitude100": 1.2, "amplitude500": 0.6, "amplitude1000": 0.3, "lownoise": 0.6, "midnoise": 0.2, "highnoise": 0.2, "amps": 0.3, "ipaddress": "192.168.1.76", "host": "rp4", "host_name": "rp4", "macaddress": "6e:37:12:08:63:e1", "systemtime": "08/26/2020 16:37:34", "endtime": "1598474254.75", "runtime": "28179.03", "starttime": "08/26/2020 08:47:54", "cpu": 48.3, "cpu_temp": "72.0", "diskusage": "40219.3 MB", "memory": 24.3, "id": "20200826203733_28ce9520-6832-4f80-b17d-f36c21fd8fc9", "temperature": "47.2", "adjtemp": "35.8", "adjtempf": "76.4", "temperaturef": "97.0", "pressure": 1010.0, "humidity": 8.3, "lux": 67.4, "proximity": 0, "oxidising": 77.9, "reducing": 184.6, "nh3": 144.7, "gasKO": "Oxidising: 77913.04 OhmsnReducing: 184625.00 OhmsnNH3: 144651.47 Ohms"}
- 14. APACHECON NA 2020 BME280 - temperature, pressure, humidity sensor LTR-559 - light and proximity sensor MICS6814 - analog gas sensor ADS1015 ADC MEMS - microphone 0.96-inch, 160 x 80 color LCD
- 15. APACHECON NA 2020 MiNiFi Java Agent ● Runs Google Net on Three Cameras ● Sends Images ● Sends classification and device data to NiFi Gateway Apache NiFi Gateway processors, validates, transforms, cleans, routes and streams events for additional processing through Apache Kafka topics.
- 17. APACHECON NA 2020 INSERT INTO global_sensor_events SELECT scada.uuid,scada.systemtime,scada.temperaturef, scada.pressure,scada.humidity,scada.lux,scada.proximity, scada.oxidising,scada.reducing,scada.nh3,scada.gasko, energy.`current`,energy.voltage,energy.`power`, energy.`total`,energy.fanstatus FROM energy, scada WHERE scada.systemtime = energy.systemtime;
- 35. APACHECON NA 2020 The code, build scripts, schemas, table DDL, Flink SQL, Kafka Connect configuration, NiFi flows, HBase tables, Kudu tables, Hive tables, HDFS directories, alerts, images, HTML, docs, links and all the goodies are here. Please fork and contribute. https://github.com/tspannhw/ApacheConAtHome2020 https://github.com/tspannhw/minifi-xaviernx https://github.com/tspannhw/minifi-jetson-nano https://github.com/tspannhw/minifi-enviroplus https://github.com/tspannhw/meetup-sensors/blob/main/flink-sql/democdf.sh https://github.com/tspannhw/FlinkSQLDemo