Mirabeau set their sights on making their offices more comfortable and healthier by translating sensor data from sensors around their building. This way Mirabeau learns about new developments in environmental sensors and data gathering tools, but also have data to develop new services in the Smart building realm.
As a first step Mirabeau started running a program with colleagues and interns that experiment with hardware, data dashboards to produce insights. My sensoring project is part of that program.
You might ask yourself: “How is this interesting to me, or my company?” Well, imagine our insights being translated into building facility services, smart building services to product a healthier workplace that can cater to the needs of its inhabitants. Operate your building more cost effective, but also more personal and flexible.
To start the Smartbuilding challenge I installed sensors through out one floor of the Kauwgomballen Fabriek (our office). This way we’re able to gather all kind of statistics about temperature, humidity, air quality and so on. But to really add value, we also measured movement and noise levels.
Step 1: Selecting the right sensors
I started to search for sensors that fit our requirements and finally decided to buy the Thunderboard sense 2 sensorkit from Silicon Labs. This board contains Bluetooth low-energy communication and a lot of sensors for a reasonable price. The availability of different Input/Output ports to add new functionality was also very convenient. This way I could connect other hardware like the PIR sensor to detect movement.
The Thunderboard sense 2 board is used for gathering temperature, humidity, CO2 particles, VOC particles, barometric pressure, office noise level (decibels), light level and occupancy information in our Digital Studio.
Soldering the different sensor parts together
For developing the software for the sensor Micrium OS was used, a realtime operating system. This allowed me to create separate running tasks inside the sensor for gathering data, cleaning data and sending it.
Ready for roll-out…
Step 2: Produce custom cases
I also needed a solid container to package our sensor-kit. This way it could be mounted safely without any change of breaking. Luckily I got some help from a colleague with 3D modelling expertise to design a nice and fitting case. To make it a bit more fun we made cases in different colors and decided to add a transparent cover to show everyone what is inside the box.
Printing the containers
Custom case with all the sensor parts
Meeting room with one of the sensors installed
Step 3: Connecting to the cloud
To collect all sensor data I decided to use a small and easy to mount Intel NUC computer that acts like a gateway to the cloud. The NUC runs Debian OS and has custom Python script installed to gather the raw sensor data. This data then is parsed, calibrated and send to the cloud. I use Microsoft Azure for this project, where the managed Azure IoT Hub service was used for data ingestion, a Cosmos DB for storing the measured data and Stream analytics for connecting IoT Hub to Cosmos DB.
Step 4: Bringing everything together
To visualize our data, my colleagues created a dashboard built on Cognizants 1Facility solution. 1Facility is a device and platform solution that integrates with existing systems and sensors to provide analytical and timely insights. Once installed and integrated, Cognizant's 1Facility Platform provides insights from multiple systems into a single dashboard, which of course received all data from the gateway.
Dashboard 1Facility software
Another option is to do some manual analysis on all data available. I asked a colleague with data science experience to support me which lead to some first insights shown in de graph below.
Relation between sound, temperature and CO2 levels.
Next Steps
It was great fun doing this project and we have already done some new IoT related projects which will be published on this blog in a while.
In this project I gained great knowledge of the Azure IoT platform and how this can be used for practical use cases. Besides this, I’ve learned how to build IoT sensors. And as a result we now have an IoT network running which is a great foundation for new innovation projects.
So what about other features? I would like to add more sensors like door-sensors and waste measurement. Doing data pattern analytics, and visualizing trends to create even more valuinsights to visitors and employees of our Digital Studio.
One of the spin-off projects is the creation of a big dashboard sharing all kind of data retrieved from our meeting rooms.
One last note: Although our setup is perhaps not suitable for a big production roll-out it definitely shows what is possible nowadays with a relative small budget. I have choosen for an approach that is more fun, appealing and open to learn how to build your own IoT setup.
Want to stay posted about IoT for your industry?
Want to know more, or stay in touch with our innovations? Follow our blog and if you have specific questions regarding this case, please get in touch with Mahadevan.ns@cognizant.com
Looking for an inspiring Computer Science Internship? Have a look here.
