Raymond Chan: Okay. So thank you for watching this abstract presentation for the DFRWS 2021. So, I’m Raymond, and today I would like to present the topic about tracing a walking path using Smart Lighting System. So, just to give you a brief introduction of what actually the Smart Lighting System is. A Smart Lighting System actually is a subsystem under the building management system, which is called the BMS. So because the BMS actually is used to monitor and control the electronic devices within the building. So in our university, because we have an Advanced Cybersecurity Lab, so we also want to explore the possibilities of using IOT systems in our campus. So that’s why we installed a Smart Lighting System in our Advanced Cybersecurity Lab. Not only to figure out are there any security or vulnerability issues, but also we would like to explore the forensic capability for the Smart Lighting System, as well.
So as you can see in this image, this is our Advanced Cybersecurity Lab. So all the light within the cybersecurity lab actually is controlled, or actually is, we installed the Smart Lighting System in one cybersecurity lab. So that when we go into the lab, the light will automatically turn on and turn off. So that actually is also saving the power bill, as well.
So, this figure is our Smart Lighting System layout. So you can see, in fact, the Smart Lighting System consists of a web application as you can see here, the web server, and this is a Zigbee gateway that is wirelessly connected to the LED light. So the LED light actually activates the Zigbee controller, and there’s a sensor to detect whenever there’s a human walking through to it. So, the application actually is not only used to monitor the situations of the LED light, but also actually you can change the behavior of the LED light as well for better management of our lab environment.
So the LED light actually is wirelessly connected using the Zigbee control. So that’s why there’s a Zigbee gateway right there to send them to the receiver and send the signal to the LED light directly.
So this is another figure to show the Smart Lighting System layout. So you can see in here we have 15 lights and in total there are seven sensors. So when someone passes by the sensor you can see the light actually will automatically turn on. So, using these figures as an example, so, if there’s someone walking through sensor 1, then you can see that lights 2 and 3 will automatically turn on. So if he walks inside to the left, then the sensor, for example, sensor 7 will then automatically detect someone passed by it and lights 6 and 12 will also automatically turn on, and so on and so forth.
So, because as you know that in the PUS diagram, I showed you that there are web applications that can retune the sensor signal. So in fact, we can also utilize that to see if we can make it as a forensic use, because it is where the information is very useful if I can know someone goes into the lab and his walking pattern, as well. So, actually we developed a Python program to reshift the information from the web applications by ourselves. And then we curate those sensors periodically to see whenever the light is now turning on or off. And we can also show then, we just put it in a CSV format to actually record the pattern and the behavior of the sensor, as well as the LED light on/off behavior.
So, actually, the results are quite impressive because when we are trying to analyze those data, we can track when someone walks into the lab, and actually which location he is currently located at. So, for example, if you’re using this prediction as an example, you can see the first indicator in the bottom left corner is that the LED light 2 actually is being turned on at 12:49 for 15 seconds. And then from time to time, you can see that LED 6 and LED 11 actually turn on later on continuously, and then LED 7 is also being turned on, as well. So using this kind of information, initially, you will only know that, “Oh, okay. That light is turned on. So if light 7 turned on after 11, then actually it’s just the pattern of the LED light.”
But, however, we can actually use this information to actually track if someone is walking to the PC. And then accordingly, so you can also see, so, if the LED light actually is being turned on and turned off, that means someone is actually walking far away from the sensor. So we also do this experiment and then we just go in to have a run through the lab. And then we found out that the LED light information is actually very useful and very accurate to detect when someone will pass by within our Advanced Cybersecurity Lab.
So, to conclude, actually, we discovered that actually we can use the sensor data of the Smart Lighting System for forensic investigations, because in the old days we only relied on the CCTV, which in most of the cases, if sometimes the CCTV is not functioning, then we might not be able to check if someone is really walking near to some locations or not. So using the Smart Lighting System sensor data can become a cost reference check. If there is also CCTV that’s right there, then we can actually adjust and synchronize the time to actually have a stronger proof of if someone is walking near a machine or near other devices.
Also, if we do not have CCTV, actually this information can also be considered as a very useful digital evidence for tracking if someone is walking within indoor locations or either outdoor locations, if possible. So in the future, basically we would like to actually set up a smart living lab on the system and we will actually test more subsystems and find out their forensic capabilities because we are actually trying to build up a so-called “Smart Campus”. So we feel that actually there will be a lot of useful sensor data that can provide small forensics capabilities, and it can also be considered as some kind of digital evidence. And that’s all from my presentation.
