Project TARAS

From January 4th when college began we started working on the EPICS project and till now the journey has been wonderful. We have learnt a lot of things and have accomplished a great deal in this past month. Here is a list of things we have done till now

1. Submitted our first proposal document highlighting our project name, its overview, the cost and a time plan.
2. Came up with the logo and a name: TARAS.
3. Made the site and a blog where we will be keeping a weekly track of our accomplishments
4. Transferred equipment from the previous batch working on a similar project thus bringing down our equipment cost by approximately Rs.6000
5. Got the data set for the accidents from 2011-2013 in the Warangal region
6. Contacted AutoLiv and Digital Ocean for support of the project

Here is the logo. It was made by Sreerag. A logo is the first step towards a larger goal. It should embody what we are doing and this shows clearly what our project aims at To protect the people, to act as shield that will warn you and help you:

We had split these four tasks among our selves so that we could complete it within our timeline.

• Aneesh and Rishabh: They were incharge of collecting data from KSVR sir. The data is an important part of our prediction algorithim since it forms the machine learning training set.
• Sreerag: He was incharge of making sure the hardware was obtained from the previous team and aquiring any additional hardware that might be neccessary.
• I was incharge of making the website fully functional and all contacting relevant organisations so that we would have a company backing us up as we moved ahead with our project

We faced a few road bumps along the way

• Getting companies: Companies or NGOs needed a working prototype that we could show before they agreed to help us. So we are moving to prototyping stage as soon as possible.
• Data intepretation: The data we recieved do not have lattitude and longitude, so we will have to generate them from the page given. This will take some time to do.

Overall, we are moving according to the timeline. Two meeting were conducted with the EPICS faculty, discussing about our progress and evaluation criteria. We understood what we have to do and how we have to go about it. We also met with our faculty mentor once and discussed the progress we have made. We had team meeting atleast once in a week so that everyone was updated with each others progress.

We have still a long way to go, but as far as I think, we are on the right track. Individually, I have started reading up about the possible machine learning algorithims we can use to tackle the problem. I have been thinking along the lines of Naive Bayes Classification but we might need a more complex classification algorithim. I have started looking into Neural networks and how far we can exploit them for our problem. I have also started reading up about accident parameters and the factors through which we can measure the intensity of an accident. We are reasearching as much as possible during the month of January so that we can get a working prototype by February end.

Here is a link to our Project Proposal. As rightfully said, A journey well begun is half done