This document discusses implementing a smart buildings solution using IoT sensors, Azure cloud services, and open source technologies. The solution aims to increase building usage satisfaction and utilization while reducing energy and cleaning costs. It generates insights from sensor data to trigger actions like notifications. It considers using Mendix for a digital twin and visualizations, InfluxDB as a high performance time series database, and the Azure IoT platform or open source TICK stack for ingesting and analyzing IoT data. The document evaluates these options and shows a demo implementation of a smart building monitoring dashboard.
Architect your app modernization journey with containers on Microsoft Azure
Report
Share
1 of 23
More Related Content
Smart building mendix azure influx / smart City / IoT
1. Smart Buildings
Actionable Data in
Mendix with
Azure & Open Source
1Smart Buildings
Gertjan van het Hof
IoT Solution Architect
IoT Enthusiast
2. Smart Building Use Case
Goals:
• Increase building usage
satisfaction.
• Increase building utilization.
• Reduce building energy
footprint.
• Reduce cleaning costs.
2Smart Buildings
4. Actionable Insights
• Notification of alerts/alarms
• Alerts/Alarms with severity level
• Functionality to adjust severity level by end-user
• Functionality to add user comments
• Rich visualizations & KPI
• Functionality to drill down to sensor readings
• Digital Twin
4Smart Buildings
5. Things
Generate data
Insights Based on
Data Generated
Actions
Based on Insights
Smart Building Solution
Sensors
LoRa
Gateway
Cloud
Platform
5Smart Buildings
6. Design Decisions
• Digital Twin in Mendix.
• Visualizations of telemetry data with Mendix Charts / Plotly.
• No telemetry data store to Mendix.
• Aggregated data and Alerts/Alarms exposed by APIs.
• High Performance (Query) Database.
• Real-time anomaly detection.
• LoRa sensors and The Things Network.
6Smart Buildings
7. Sensor used in this use case
• LoRa Alliance Certified
• Temperature
• Humidity
• Light
• Occupancy (body sense)
• NFC for easy configuration
• Over the air configuration
• Size : 86x86x26mm
• Accuracy: ± 0.5°C, ±2%rh
• Resolution: 0.1°C, 0.1%rh
• Approx. range: 8km*
• Battery life: 5 years**
• US902-928, EU863-870, AS923, AU915-928, KR920-923
• 2 x 3.6V AA lithium battery
LoRa ERS Eye from ELSYS
7Smart Buildings
9. What type of database for sensor data do we need?
• The data is (mostly) recorded as a
new entry.
• The data usually arrives in time
order.
• Time is a primary axis (time-
intervals can be either regular or
irregular).
• High number of sensor readings.
9Smart Buildings
11. InfluxDB vs MongoDB
• InfluxDB outperformed MongoDB by 2.4x when it
came to data ingestion.
• InfluxDB outperformed MongoDB by delivering 20x
better compression.
• InfluxDB outperformed MongoDB by delivering 5.7x
better query performance.
11Smart Buildings
12. Azure IoT Solution
12Smart Buildings
IoT Hub Streaming
Analytics
CosmosDB
Function
Apps
API
Gateway
Machine
Learning
Anomaly Detection &
Data Aggregation
Cloud Entry Point Alerts & Data
Telemetry Data Store API’sDevice Management
22. Takeaways
• Community Edition of most Open Sources products is single node.
• Implementation speed with Azure Services is higher compared to Open Source alternative.
• Azure ML is closed environment. Alternative is Azure Data Bricks, but this is a complete
different architecture and much more complex.
• With Influxdata TICK stack + Loud ML a cost effective solution can be build.
• For advanced ML Use Cases use TensorFlow with Jupiter Notebook.
• Mendix can be used for making data actionable but complex visualizations can be
problematic.
22Smart Buildings
Editor's Notes
ASB = Anti Social Behaviour
Internet of Things is het aansluiten van fysieke objecten aan het internet
Daarmee vragen we de status op en monitoren we het gebruik
Op basis van deze data trekken we conclusies
Die resulteteren in acties in het ERP system
Of direct ingrijpen op het aparaat zelf