This is a use case study first published on Marktplatz I4.0 in German. qbee AS is a member of the industry 4.0 initiative Marktplatz I4.0 e.V. It shows how remote device management and remote access help facility management companies to monitor & operate their smart buildings more efficiently and environmentally friendly.
How does qbee.io help with smart buildings?
A huge facility management company has risen to the challenge to operate their property portfolio in a more efficient and environmentally friendly way. They want to turn their facilities into smart buildings with the help of wireless sensors. The goal of this project is to get a detailed overview over the indoor climate in both new and old office buildings. Based on the analysis of the indoor climate relevant energy efficiency measures will be taken to lower the energy footprint of each building. The buildings are located over the whole country and all sensors as well as the receiving gateways need to be installable by each building manager in order to save installation costs. The operation of the application software as well as the gateways (the edge gateways are operated with an embedded Linux operating system) are controlled centrally.
The project was delivered by Greenflow AS. It consists of the following components:
- Cloud IoT platform Tingsense
- Multi sensor system & 4G edge gateways (an industrial Raspberry Pi) with wireless measurements of CO2, temperature, humidity, light and noise level
- Embedded Linux device management and remote access software qbee
A Service Level Agreement is only as good as the weakest component in the technology stack
It became obvious very quickly that the planned service level agreement (SLA) could only be reached when all gateways in all buildings would be actively monitored. A smart remote device management system will be able to reduce errors and downtime. That was the reason qbee became a central part of the smart building architecture providing both remote access as well as device management. In order to be independent of the buildings’ network it was decided to rely on 4G cellular connectivity for gateway to server communication. This allows to exclude the smart building’s own network as a source of error and less communication between the central monitoring and each building has to occur.
An IoT platform for smart building and facility management that monitors indoor climate
All smart buildings deliver their sensor data to the IoT visualization platform Tingesense. Within Tingsense different climate zones can be defined and wireless sensor data can be assigned to areas, floors and buildings. Thus, a bottom up aggregation approach allows very granular insight into indoor climate conditions which makes it ideal for analysis. The platform contains a PMV ( predictive mean vote) / PPD (percentage people dissatisfied) model. From this model the control parameters for the climate control are derived. The PMV/PPD model is based on very extensive studies of human perception of temperature in relation to humidity, clothes, metabolic activity and HVAC air flow in the areas. In the image on the left a detailed dashboard visualization of a building with their respective PMV/PPD points is shown. Deviations from the ideal combination of temperature and humidity are recorded and a “satisfaction index” is calculated and visualized on the dashboard. This only calculates working hours and excludes when setpoints are lowered during the night or weekend. Because this satisfaction index is calculated with measured and defined model parameters the energy consumption of buildings can be compared. But now you can also correlate with the quality index. This allows to understand if the energy efficiency is achieved by compromising indoor climate quality or through smart regulation of the components in the HVAC. This allows to compare energy consumption across buildings and this time it is not the most penny-pinching facility manager that wins but the one that combines energy efficiency with good indoor climate. This is a win-win situation making the building more environmentally friendly, more cost efficient and also more pleasurable to work in.
In addition to aggregated analyses of areas or floors it is also possible to get detailed insights on a per sensor level to see irregularities, analyze patterns and generate additional insight. Below is a Tingsense IoT platform dashboard visualizing a single sensor:
Why do you need IoT edge device management and remote access?
A major selling point of the system is that all edge gateways and wireless sensors can be installed very easily by the respective facility manager. Then they need to operate reliably over the whole lifecycle of this smart building or facility management system. That is why qbee AS was contracted as a partner to automate the edge gateways completely with their smart remote device management platform. In addition the possibility to create a simple means of having secure remote VPN access through qbee’s built-in remote access feature was a key value add. With this device management solution a complete system can be delivered to each building and the edge gateway can automatically connect to the IoT platform from qbee as well as to the cloud backend for the IoT visualization platform Tingsense.
qbee is an IoT device management platform for embedded Linux edge devices. This platform works by installing a small agent software on the edge device which then securely connects to the qbee backend in the cloud. A unique feature of qbee is that it runs on any Linux device. This allows to connect also older and existing other units into the IoT device management platform. Through the pre-compiled agent approach there is no need for software development and the whole IoT solution is offered as a hosted solution. Greenflow AS was able to test the gateways and sensors both in the lab but also on premise at the customer site from day one. This helped to show the customer new functionality very quickly and new possibilities were discovered through this active collaboration from day one.
The possibility to receive data from real buildings early in the project was very helpful. At the same time the software was constantly updated on the gateways through the OTA software update functionality of qbee. This allowed Greenflow to constantly show the customer progress and milestones. This established a very good trust based relationship especially at the beginning of the project. It was also possible to introduce new impulses into the agile software development process based on the customer’s feedback. With the qbee remote device management solution frequent software updates worked flawlessly and the included VPN remote access opened various possibilities to access and debug the devices remotely.
What features does an edge device management solution cover?
In this project qbee is used to cover the following areas:
- OTA software updates of the OS and application software
- Configuration management to configure the Linux operating system
- Parametrization of the application software (MQTT data broker)
- Remote access for analysis and debugging
- Monitoring of the edge device performance
- CVE vulnerability analysis and continuous monitoring
- Monitoring of edge gateway functionality
- Watchdog for the 4G cellular USB modem
The key driver for the selection of qbee in this use case was the remote device access functionality. But in the end most of qbee’s features were used and the customer was positively surprised how many value adding functions the tool offers. That is why we often refer to qbee as the “swiss army knife for embedded Linux systems”.
Important software updates are automatically deployed on the remote edge gateways. This applies to the application software as well as for the packages of the operating system. This over-the-air (OTA) software update is an important aspect for the operation of the system. In addition, a CVE vulnerability analysis continuously compares all package libraries with entries in the NIST CVE database. This allows to pro-actively identify libraries that show vulnerabilities in addition to being able to update those as soon as a patch is available. With the built-in state-based configuration engine it is possible to handle all the Linux OS housekeeping. This can be the user or password management, NTP server configuration, ssh key rotation but also to configure the firewall such that all ports are closed. Since all remote device access works through the qbee built-in VPN no ports need to be open and accessible from the outside. Through a pull-based architecture the agent can initiate communication through all firewalls, NATs or in this case cellular networks. No IP address of the device or building needs to be known. The agent will automatically connect to the qbee cloud solution. All this happens just by installing and bootstrapping a pre-compiled agent. No development is needed. Immediately after the installation both remote ssh access over the web interface as well as with the included qbee-connect desktop application for Windows, MacOS and Linux can be achieved. It is possible to access any port or service (http, https, Node-Red, VNC, SCADA Systeme) and even ssh port forwarding of other closed devices in the local network is possible through this secure solution (e.g. legacy HVAC devices with a local web control interface on port 80 or 443).
Both for the development and operational phase it is very convenient to be able to work with metric data that gets collected from the devices and that is visualized on a dashboard. Amongst other things, CPU load, memory consumption and consumed bandwidth are visualized. It is possible to use different timescales or zoom. Application memory leaks are clearly visible over time and log creeping is also easily to identify. If a device is offline this is immediately displayed and a warning can be issued.
The gateway solution consists of an industrial Raspberry Pi from Techbase with built-in cellular USB modem for 4G connectivity. The proprietary sensors are also connected through a radio module utilizing the USB bus of the device. During the project it was noticed that sometimes a sim card re-configuration from the cellular connectivity provider caused the Linux network stack to wrongly assume connectivity was still present while the radio had lost it. But through the smart watchdog functionality of qbee it was possible to correct this problem very late in the project. The watchdog was simply configured through the configuration management engine and then all devices started to work with the watchdog.
The application on the gateway communicated through MQTT with the Tingsense cloud. Through the parametrization of the application software with templating it was possible for Greenflow to control different MQTT broker settings on a group or device level. This allows to use the same solution for multiple customers, or a system operator can provide services to monitor different fleets of devices for different customers through one system and dashboard.
We are happy to consult you how embedded device management and remote device access can help your project. Focus on your core business area and core competency. We will help you with the device management and device monitoring over the complete lifecycle of the devices.