How The Netherlands blew up €68mln

Document created by on Jan 13, 2017Last modified by on Jan 13, 2017
Version 2Show Document
  • View in full screen mode

How The Netherlands blew up €68mln

And we used €12 worth of hardware, InfluxDB & ArcGIS to track it



Every New Years' eve masses of fireworks are lit in The Netherlands. I'd like to take this opportunity to explore how masses of cheaply produced IoT sensors can not only greatly enhance measurement quality and sample size, but how easy it is to integrate these in an ArcGIS for Enterprise environment. Techniques used: Javascript, Webappbuilder, Portal, Geoevent Server, Geoevent Server SDK + Java, InfluxDB, C#.



No doubt over the past year everyone’s been buzzing with “IoT” and it’s real-world applications. How everything’s going to be better (or worse) because of it, and how it’s going to solve many socio-economic issues.


As at any tech consulting/integration firm, a lot of Ordina clients often close down during winter recess, which gave me time to approach our innovation unit who, amongst my own unit, are working on IoT.


This year though, we had a business case. We wanted to “enhance” the Dutch Governments’ initiative on crowdsourced fine particle/air quality measurements. Especially during new years’ eve this is a huge deal: €68mln worth of fireworks get blown into the air each year, which is subject of a lot of national and local debate. Measurements like these are already being taken using highly specialized equipment for many years (also around roads), however we wanted to optimize and make this cheaper.


Using a Shinyei PPD42 particle sensor (€8) and a ESP8266 WiFi/SOC module (€4) we managed to build a surprisingly precise low-power sensor. We distributed 30 of these to Ordina colleagues, and asked them to hang these around their houses. Using the same sensor, but with a LoRA module rather than WiFi (LoRA = an IoT platform with nationwide coverage in The Netherlands; on the old UHF frequencies; stand-alone long range sensors can last for half a year on a battery), we even managed to place a few “in the wild” being fed with a battery.


Due to customer constraints a free and open source time-series database was used; InfluxDB. Time Series databases specialize in real-time monitoring (near-analog levels of accuracy) of data and measurements, with time being the defining factor. In order to connect to ArcGIS, Geoevent processor 10.5 was used to integrate all data with GIS. The result was a live viewer which also was able to show historical data with a highly versatile data model. It was viewed during new years’ eve by many Ordina Employees (one even tracked his own sensor through GIS on a Nintendo 3DS even…), and the results (how GIS strengthened this solution) were shared with said government agency – who hadn’t considered using GIS in the past. The 32 sensors tracked now could just as easily have been 3200 sensors!


Best thing? Everything from data to GIS was built by one man in a single week. IoT & GIS: Very Powerful, yet not a big deal. As for the results: the €68mln worth of fireworks were causing up to 40 times more fine particles in the air than regular. Which is only 4 times as much than the surprisingly trackable morning rush hour in some parts of the country.


This session shows the steps taken, how InfluxDB works, how we integrated it in ArcGIS with Geoevent Processor, and how the final viewer was built, and why IoT is so easy and cheap to implement.

8 people found this helpful