8th Semester Research Internship
Once a virtuous red panda said “If you only do what you can do, you will never be more than you are now”. This quote exemplifies my Internship journey which I did at School of Cybersecurity, Georgia Institute of Technology under the guidance of my mentor Moses J Ike. Before proceeding further a little bit of introduction. Bonjour I am Asfahan Shah a final year undergraduate student pursuing computer science at Bennett University. Now with introduction out of the way, lets get to the main dish and answer some sizzling questions like what I did in in internship, what was the objective of the internship and what I learnt during internship.
Before going into detail about answering these questions, lets take a moment to answer how did I got the internship.
Well see my university had a collaboration with Georgia Tech where each year top meritorious students get a chance to conduct research at Georgia Tech. And due to my excellent academic performance (yes there is still value to getting more than 7.5 CGPA), I was able to get the internship.
Now what was the internship about you may ask
The internship was about solving the problems of cybersecurity using Machine Learning. Now you may ask what domain of cybersecurity and what problems ? In the internship, my job was to devise solutions that can detect anomaly in cyber physical systems.
What is cyber physical systems? is it like a ChatGPT, I hear you ask.
Cyber physical system is a system in which a mechanism is controlled by computer based programs. Like your electric grids, autonomous vehicles, industrial processes and so on.
So yeah, its cool and all but now what is the problem here
the problem lies with the fact since these system are controlled by a computer algorithm, there is a high chance some malicious agent may launch an attack on these CPS (cyber physical systems). And these attacks can cause severe damage. Like on a critical CPS like power grid a small attack can possibly cause power outages.
Ok cool but what did you actually work on, you may ask
Well in this internship I got the opportunity to work on not one but on two projects. These projects were Third Eye and ICSBRIDGE. Both of these projects are on the same principle of anomaly detection in CPS but Third Eye deals with power grids while ICSBRIDGE deals with industrial process
A little bit of elaboration regarding these projects
ICSBRIDGE is a system that utilizes something called physics informed neural network. This is just a fancy term for saying that the model utilizes a partial differential equation (PDE) in its loss function. Now you may say PDE, is there any way to circumvent it. Well no see PDEs are th heart of physics informed neural network. But hey one silver lining or you can AI magic is that we do not need to solve this PDE, instead AI will solve it for us.
Now Third Eye is again used for CPS but for power grids. More concretely we are using something called HMI images to train the autoencoder model in order to detect anomalies.
Hey what is HMI image ?
HMI is human machine interface image, or an image which our human operator will see on the control monitor.
Now coming what I actually achieved during the internship
Well, I am really happy to say that ICSBRIDGE project has been converted into a research paper, which hipefully by the time this blofg is published will be summitted to a research conference. Now what about Third Eye. Well it is completed partially in the sense that yeah project works and all but it still needs a little bit of fine tuning before we can convert it into a research paper. Hopefully during next two-three months we will be able to achieve it.
Conclusion
This internship was something which I never expected, if you would hav told me that for next 6 months you would work in cybersecurity domain I would have laughed at you. but now after completing this internship, it feels rewarding. Rewarding to not only complete an intensive 6 months research, but also to push myself beyond my comfort topic and innovate to get the solution.
Overall an experience, which is something which I will never forget and in the end I would advice everyone to follow the advice of virtuous red panda
