Discovering a Malfunction In the Home Heating System With Mainflux IoT Platform
IoT solution for monitoring a home heating system using Mainflux IoT platform with real-time data collection visible in one single application
In this demo made by Boris Bokun founder and CEO of our partner company strictit, the goal was to create an IoT solution for monitoring a home heating system using Mainflux IoT platform.
Boris and his company have more than 20 years of experience in creating industrial software solutions for Quality Management and industrial automation deployed mostly in German SME as well in industrial facilities on 4 continents of the world.
With the rise of Industry 4.0 concept and internet of things technology which can provide significant enhancements to existing software solutions that have been used in industry so far, Boris naturally started exploring how he can improve his projects and company offering.
He experimented with a lot with different systems and at the end chose Mainflux IoT platform.
As the first demo, he made a solution which provides him the possibility to collect real-time data of his home heating system and to have this data visible in one single application. To explore additional possibilities of Mainflux IoT platform he used a Hololens which creates an industry setting feel to the demo.
Demo resulted unexpectedly in discovering a malfunction in the heating system which caused no small loss of energy.
BACKGROUND
Home heating system:
- Bosch SOLID 2000B solid fuel boiler
- Brenn 2000 pellet burner
- Hot Water Tank
- Heating Buffer
- Solar hot water panels
- A microprocessor aided double differential thermostat
- Tech ST-717 controller controlling the boiler (ignition, feeder…)
- Tech ST-505 Internet module for Tech controllers (thanks go to Bosch Termotehnika d.o.o Belgrade, who kindly borrowed to Boris the module for testing)
Software:
Data gathering:
Windows Service created in Delphi Rio (https://www.embarcadero.com/products/delphi) for acquiring data from ST-505 module and publishing to Mainflux IOT platform
Dashboard
- Application — Front-End developed in Angular
- Back-End created in Spring-Boot
HoloLens
An application created in Unity 2017.4.18f, compiled in Visual Studio 2017 Community Edition.
Boris Bokun, CEO of strictit, with Microsoft HoloLens
Computer equipment:
At home
- Windows Server running the Windows Service
- HoloLens (thanks go to our big friend AJ)
In the cloud:
- Hyper-V Server 2012 R2 running:
- CentOS 7.5.1804 guest for Mainflux IOT platform (version 0.8.0, with deployed InfluxDB-writer)
- Windows OS guest for the Dashboard application
CHALLENGE
Boris’ heating and domestic hot water preparation system combine a Bosch Brenn 2000 pellet burner and Two Solar Panels installed on the roof. In the winter, the domestic hot water is prepared by almost 100% from pellets. In spring the solar panels start to deliver the energy for the hot water, until late autumn. Of course, this depends on the sunshine.
A microprocessor aided double differential thermostat regulates the heating of the domestic hot water from the two heating sources.
In the original configuration without Mainflux, the Tech ST-505 ethernet module sends data every few seconds to a website which is run by the manufacturer of the controller, providing very basic and insufficient analysis.
That was the problem because the idea was to have a solution which would be more flexible and gives the possibility to perform data analysis in a more complex way.
SOLUTION
The Windows Service on the home Server polls for values from the ST-505 Internet module on a regular basis via HTTP. The data is published to Mainflux. Additionally, Microsoft HoloLens and the Dashboard application are subscribed to Mainflux MQTT.
Values which are collected from the equipment are
- Central heating temperature — goal value, the actual value
- Hot Water temperature — goal value, the actual value
- Status of boiler — on/off, ignition
- Status of the feeder — on/off
- Speed of burner fan
How does it work?
In Mainflux a Thing is configured with two channels — one channel for JSON data for actualizing the Front-End(s) and one channel for logging purpose — in SenML JSON Format.
The Windows Service on the home Server polls for values from the ST-505 Internet module on a regular basis via HTTP. The data is decoded and published to the two Mainflux channels via MQTT into the cloud. The applications for data presentation (Dashboard Widgets, HoloLens) are subscribed to the first channel, get the data from the broker, unpack it, and visualize as needed.
The data on the second channel — which is in SenML format, is automatically caught by Mainflux IoT platform, and the mainflux-normalizer converts raw SenML messages and forward them to the Mainflux message broker to store those messages into the database through InfluxDB writer.
The Dashboard has also widgets for showing Grafana dashboards. The dashboards were “clicked together” in the Grafana Dashboard Editor.
Benefits
As he got the data in a performant and practical way, Boris and his team from “strictit” were able to create an application for visualization of real-time data, which is subscribed to Mainflux MQTT.
Mainflux’s ability to automatically store data in time series InfluxDB database provided the evaluation of historical data by using Grafana. The dashboards were “clicked together” in the Grafana Dashboard Editor and put as widgets into their application.
The following Dashboard shows real-time data and historical data (Hot water history graph and central heating history graph from Grafana)
Thanks to our friend AJ, who gave the idea to experiment with AR technologies, the challenge of the presentation of real-time data with HoloLens started.
Afterward, a compared evaluation data of Hot water and central heating temperatures were created and suddenly a discrepancy in the heating system was identified.
Last summer (and we had a very hot summer), the need for warm water was completely covered by the solar hot water panels.
But, somehow, Boris had the feeling that something is wrong — as the pipes of the heating branch were also getting warmer when the solar panels heated the hot water in the tank. Boris’ initial suspicion was that a one-way check valve, which separates the heat exchanger from the pellet boiler branch is defect and water in the boiler branch is heated up. He now could prove his suspicion with the evaluated data (see picture — the crosshair shows the situation — both the water tank temperature — yellow — line and the central heating temperature — green — line are raising). He called the maintenance service which replaced the one-way valve, and now the system works as expected.
Additionally, another value of the collected data is the possibility to do a prediction of pellet consumption by summing up the uptime of the pellet boiler and pellet feeder, which will help to finally be able to calculate the needs of pallets for the next winter rather than guessing.
VIDEO
strictit
The Company strictit was founded in 2015 after a successful spin-off from the company Pragmatic-IT which was founded in 2003 and sold to a company from the Phoenix group — one of the leading pharmacy wholesale in Europe-
the strictit team also works for more than 15 years as outsourcers for one of the best software producer of Quality Management Software and industrial automation in Germany for the automotive industry.
website: http://www.strictit.com/
Mainflux
Mainflux is a technology company offering a full-stack open-source, patent-free IoT Platform recognized by Linux Foundation and O’Reilly Media with software and hardware consulting services based on extensive working experience in fortune 500 companies.
website: https://www.mainflux.com/
GitHub: https://github.com/Mainflux/mainflux
Mainflux Facts:
200,000 + Docker Hub downloads
136 Forks on the Github
Member of Linux Foundation EdgeX Foundry (in TSC) and LF Edge with companies like AT&T, Dell, Baidu, HP, Ericsson, Huawei, Nokia, Red Hat, Samsung, Intel, and IBM.
