The data collected should include temperature, humidity, co2 levels. Additional sensors that could be used is a motion sensor to automatically detect how many people enter a room, but that is not mandatory. Air quality should be our focus on the project. Light levels could also be measured and then we need to consider light colour as well. Correct placement of the sensors is a key thing for quality data. We also need to observe the amount of people in the room and how long they are present as they directly correlate to the air quality in the room.
The digital twin model would have real time data from a building that would automatically display the data on the virtual counterpart. Digital twins integrate IoT sensors, AI and machine learning to create the living digital simulation. This model would use the 3D-model of the building and visualize the data in it. Turku GameLab should have the 3D-model for the building that we can use. This requires large amounts of data for the machine learning, as you can not train AI without data. The visualization is also a key part of the project.
Data security is also an issue we should consider on the project. Environmental data could be interpreted as personal data in some cases, so security is required to keep data private. Some officials say that sensing in spaces is under GDPR law, motion sensors could be considered as cameras and those would require people to opt in for the data gathering. Normal air quality sensing is not directly related to persons, but it can be combined with other data that could link it to a person so it would fall under GDPR. Testing in classrooms might require permissions from Kiinteistöpalvelut, and we need to know if we need to inform teachers etc. about testing in a classroom during a lecture.
For data security we could do cybersecurity analysing and testing for IoT systems. Good security is a must for smart buildings. Security planning and risk analysing would need to be done and testing the system on how secure it is.
For the coding languages we can use what ever we want but keeping it simple is a good choice. If we use Unity for the 3D-modeling it uses C#. Data will be gathered to Azure Cloud but that does not restrict code usage, so keeping it in Python or C# is a good idea.
For data transfer the project owner wishes to use a testing network, that would require permissions from other persons not involved directly in the project. If that is not possible, we can use mobile networks for data transfer.
Research on machine learning is needed to use it effectively in the project to build the behaviour models. We also need to find research data on room air quality and use that as a base for planning our own monitoring and testing. Experimenting with the sensors on how they behave is also important to gather quality data. There are several key figures that need to be considered for the data so it can be used in the project effectively. For example, the amount of people affects the levels of CO2 directly over time.
There has been previous work done on the project, so we do not have to start from scratch. The project owner will provide us access to the work that has been done previously. Parts of the project need to be developed at the same time as they are dependant on each other, realizing roadblocks and dependencies is key for the project to move forward smoothly.
Clear planning for the project is important. If we need to order certain sensors it will take time to acquire them for the project. Regular face to face meetings with the project owner is a good idea so we can keep everyone onboard and plan the project forward accordingly.
No comments:
Post a Comment