O slideshow foi denunciado.
Utilizamos seu perfil e dados de atividades no LinkedIn para personalizar e exibir anúncios mais relevantes. Altere suas preferências de anúncios quando desejar.

Pushing it to the edge in IoT

25 visualizações

Publicada em

Where is the edge in IoT and how much can you do there? Data collection? Analytics? I’ll show you how to build and deploy an embedded IoT edge platform that can do data collection, analytics, dashboarding and much more. All using Open Source.

As IoT deployments move forward, the need to collect, analyze, and respond to data further out on the edge becomes a critical factor in the success – or failure – of any IoT project. Network bandwidth costs may be dropping, and storage is cheaper than ever, but at IoT scale, these costs can still quickly overrun a project’s budget and ultimately doom it to failure.

The more you centralize your data collection and storage, the higher these costs become. Edge data collection and analysis can dramatically lower these costs, plus decrease the time to react to critical sensor data. With most data platforms, it simply isn’t practical, or even possible, to push collection AND analytics to the edge. In this talk I’ll show how I’ve done exactly this with a combination of open source hardware – Pine64 – and open source software – InfluxDB – to build a practical, efficient and scalable data collection and analysis gateway device for IoT deployments. The edge is where the data is, so the edge is where the data collection and analytics needs to be.

Publicada em: Software
  • Seja o primeiro a comentar

  • Seja a primeira pessoa a gostar disto

Pushing it to the edge in IoT

  1. 1. @davidgsIoT David G. Simmons / @davidgsIoT PushingIoTDatatotheEdge
  2. 2. © 2019 InfluxData. All rights reserved. @davidgsIoT Where To collect your IoT Data? •Distributed collection vs. Centralized collection •How is the data used •Who uses the data
  3. 3. © 2019 InfluxData. All rights reserved. @davidgsIoT Collect it in the cloud •All data flows directly to the cloud •Requires highly available network –Low latency –No down-time •Analysis and visualization from anywhere
  4. 4. © 2019 InfluxData. All rights reserved. @davidgsIoT Collect it at the edge •Collect close to the sensor •Unreliable back-haul network •Local processing and analysis
  5. 5. © 2019 InfluxData. All rights reserved. @davidgsIoT Distributed data collection •Data collected at multiple collections points •Remote collection points feed back-end system of record •Distributes data collection load •More tolerant of network outages, etc.
  6. 6. © 2019 InfluxData. All rights reserved. @davidgsIoT Data Layer Architecture •Data collected at the edge, where it’s generated •Edge collectors also capable of analysis •Edge collectors handle local event, etc. •Down-sampled data forwarded to backend on a network-available basis –Lower network costs –More fault tolerant
  7. 7. © 2019 InfluxData. All rights reserved. @davidgsIoT A Rational IoT Architecture InfluxData • Telegraf • Data Collection • InfluxDB • Short-term storage • Long-term storage • Kapacitor • Local Alerts • System-wide alerts • Chronograf • Dashboards
  8. 8. © 2019 InfluxData. All rights reserved. @davidgsIoT How does InfluxData help? • Extremely efficient data collection • High-volume data collection • IoT generates huge volumes of data very quickly • Being able to ingest, analyze and query that data is key to IoT success • Ease of Deployment • Easy to deploy InfluxDB and the TICK stack for data collection, analysis and action • Very low time to value – Time To Awesome™ • Dashboards and visualization • Easy to build useful, easy to read dashboards.
  9. 9. © 2019 InfluxData. All rights reserved. @davidgsIoT Dashboards Everywhere
  10. 10. © 2019 InfluxData. All rights reserved. @davidgsIoT Monitor the Platform
  11. 11. © 2019 InfluxData. All rights reserved. @davidgsIoT What is flux? Flux is a functional data scripting and query language • Written to be: – Useable – Readable – Composeable – Testable – Contributable – Shareable • JavaScript-esque • MIT Licensed
  12. 12. © 2019 InfluxData. All rights reserved. @davidgsIoT Arduino Native #include <InfluxDb.h> #define INFLUXDB_HOST “myhost.com” #define BATCH_SIZE 10 Influxdb influx(INFLUXDB_HOST); void setup() { // our debugging output Serial.begin(115200); influx.setBucket("telegraf"); influx.setVersion(2); influx.setOrg("influxdata"); influx.setPort(9999); influx.setToken("3hbWcTRvQ5HtgGWlOmDPVX2fjUDgH2s3kY- kaIbvk53EDQkHaKfzhXfPlXsj82qttHOin1NBxX7ZBY4Rm4AydA=="); }
  13. 13. © 2019 InfluxData. All rights reserved. @davidgsIoT Batch Writing void loop() { if (readPMSdata(&pmsSerial)) { if(batchCount >= BATCH_SIZE){ influx.write(); batchCount = 0; } InfluxData row("particulate"); row.addTag("sensor", "pm_sensor"); row.addValue("particles_10µm", data.particles_10um); row.addValue("particles_25µm", data.particles_25um); row.addValue("particles_50µm", data.particles_50um); row.addValue("particles_100µm", data.particles_100um); influx.prepare(row); batchCount += 1; }
  14. 14. © 2019 InfluxData. All rights reserved. @davidgsIoT InfluxData as the IoT Data Platform • What is IoT Data? – sensor@time – that’s time series data! • IoT data MUST be – Timely – ingestion rates and query efficiency is key – Accurate – data integrity and platform reliability is important – Actionable – data visualization, anomaly detection & alerting are essential • IoT deployments are struggling to find efficient, scalable, data platform that meets all of these criteria • The InfluxData Platform meets them all
  15. 15. @davidgsIoT THANKYOU